Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
Chapter 4:
Glacier mapping and monitoring based on
spectral data
Andreas Kääb1, Tobias Bolch2, Kimberly Casey1, Torborg Heid1, Jeff Kargel3, Greg Leonard3,
Frank Paul2, Bruce Raup4
1
(Department of Geosciences, University of Oslo, PO Box 1047 Blindern, 0316 Oslo,
Norway, kaeaeb@geo.uio.no)
2
(Department of Geography, University of Zurich, Winterthurerstrasse 190, 8057 Zurich,
Switzerland, frank.paul@geo.uzh.ch, tobias.bolch@geo.uzh.ch)
3
(Department of Hydrology and Water Resources, University of Arizona, PO Box 210011,
Tucson AZ 85721-0011, USA, kargel@hwr.arizona.edu)
4
(National Snow and Ice Data Center, University of Colorado, 449 UCB, Boulder
Colorado 80309-0449, USA, braup@nsidc.org)
1
Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
Table of Contents
Abstract ........................................................................................................................................... 4
4.1
Introduction .......................................................................................................................... 4
4.2
Image preprocessing ............................................................................................................ 5
4.2.1
Radiometric preprocessing ........................................................................................... 5
4.2.2
Geometric preprocessing .............................................................................................. 6
4.3
Multispectral methods .......................................................................................................... 8
4.3.1
Spectral response of typical glacial surfaces ................................................................ 8
Snow ....................................................................................................................................... 8
Ice ............................................................................................................................................ 8
Rock ........................................................................................................................................ 8
Water ....................................................................................................................................... 9
Vegetation ............................................................................................................................... 9
4.3.2
Principles of classification approaches ....................................................................... 10
Hard versus soft classification .............................................................................................. 10
Manual, supervised and unsupervised classification ............................................................ 11
Parametric versus non-parametric classification .................................................................. 11
Spatial and spectral segmentation ......................................................................................... 11
Subpixel classification .......................................................................................................... 11
Combinations and others....................................................................................................... 12
4.3.3
Image processing techniques for glaciers ................................................................... 12
Manual delineation................................................................................................................ 12
False color composites .......................................................................................................... 13
Calculation of reflectance ..................................................................................................... 13
Spectral transforms ............................................................................................................... 14
Image algebra and segmentation........................................................................................... 16
Unsupervised classification .................................................................................................. 17
Supervised classification ....................................................................................................... 17
Artificial Neural Networks (ANN) ....................................................................................... 18
Combinations ........................................................................................................................ 18
4.3.4
Post-processing and embedding in a GIS workflow .................................................. 18
4.4
Mapping of debris-covered ice .......................................................................................... 19
4.5
Thermal imaging ................................................................................................................ 21
4.6
Microwave/SAR methods .................................................................................................. 23
4.7
Spectral change detection; multitemporal data merging .................................................... 28
4.7.1
Overview .................................................................................................................... 28
4.7.2
Image Change Evaluation by Subtraction of Multispectral Anniversary Pairs
(ICESMAP)............................................................................................................................... 31
4.8
Ice flow .............................................................................................................................. 37
4.8.1
Image choice and pre-processing for image matching ............................................... 38
4.8.2
Image matching techniques ........................................................................................ 39
4.8.3
Post-processing and analysis ...................................................................................... 40
4.8.4
Accuracy ..................................................................................................................... 41
4.8.5
SAR offset tracking and interferometry ..................................................................... 42
4.9
Challenges, conclusions and perspectives ......................................................................... 45
2
Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
3
Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
Abstract
Spectral, and not least multi-spectral satellite data represent the backbone of spaceborne glacier
mapping and monitoring, and have allowed for substantial progress of global glacier
observations in recent years. In this chapter we give an overview of the information contained in,
for the most part, ASTER or Landsat type spectral data, and of methods to qualitatively and
quantitatively analyze this information for glacier studies. Besides multispectral techniques
based on the visible, and near- and short-wave infrared sections of the spectrum, we also shortly
discuss methods based on thermal and radar data, with special emphasis on the mapping of
debris-covered glacier ice. A further focus is on spectral change detection techniques applied to
multitemporal data over glaciers, with special attention to a novel image differencing technique.
Finally, we give an overview of satellite-image based measurement of glacier flow, a group of
methods that we view as a category of spectrally-based change detection.
4.1
Introduction
This chapter gives an overview of mono- and multitemporal approaches to map and monitor
glaciers from space based on spectral data, of the kind of information that can be retrieved, and
of the information characteristics and accuracy obtained. Specifically, we consider in this
chapter:
•
image segmentation and classification of mono- and multispectral data for mapping
glacier outlines and glacier zones;
•
spectral detection of changes in glacier boundaries and zones over time using
multitemporal data;
•
geometric detection of ice displacements (ice flow) using repeat data.
These three groups of glacier-related information represent, together with the production of
glacier topography and the detection of elevation changes over time, the backbone of glacier
observation from space. Geometric analysis of stereo data for derivation of digital elevation
models (DEMs) is not considered in this chapter (see Chapter 5). Further, the main focus of this
chapter is on the visible to short-wave infrared section of the electromagnetic spectrum. Thermal
and microwave data are only covered in brief.
In general, spectral remote sensing of glaciers is greatly facilitated by the large difference in
snow and ice reflectance from high values in the visible (VIS) and near infrared (NIR)
(combined: VNIR) to low values in the short-wave infrared (SWIR) section of the spectrum.
This specific feature in the optical spectrum represents the core for most multi-spectral glacier
mapping. On the other hand, spectral remote sensing of glaciers is complicated by the fact that
4
Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
glacier areas other than snow, firn and clean ice can be spectrally very similar to the surrounding
of a glacier (e.g. debris-covered ice). In radar images, the backscatter contrast between ice and
snow, and other terrain can be small and temporally unstable, so that a corresponding image
segmentation is usually not straight forward. Furthermore, a spectrally defined glacier will in
many cases not be equivalent to its definition in terms of an ice-dynamical or mass-balance
system (e.g. snow patches attached to a glacier, outlet glaciers draining from one single ice cap).
Finally, whereas many image analysis methods result in sharp boundaries between different
classes, nature may contain gradual transitions that are then not adequately captured in the digital
data (e.g. gradual change of debris-cover percentage). These major opportunities and problems
characterize most methods and attempts at glacier mapping and monitoring using spectral data.
4.2
Image preprocessing
In most cases, the remotely sensed data used for glacier mapping and monitoring will have to
undergo radiometric and geometric pre-processing. Radiometric pre-processing concerns pixel
values, while geometric pre-processing concerns pixel locations. Radiometric pre-processing can
include corrections for instrumental gain and offset, illumination conditions, as well as
atmospheric absorption and scattering. Geometric pre-processing can include geo-location,
registration and or re-projection of image geometries.
4.2.1 Radiometric preprocessing
A first step in radiometric preprocessing is the correction of systematic differences and variations
in the sensitivity of detectors. In pushbroom imagery, such as ASTER, these sensitivity
differences lead in raw imagery to along-track stripes. The necessary radiometric correction
coefficients (gain and offset, or higher-order polynomials) are provided with the raw image data
(ASTER Level 1A) or already applied (Level 1B and up). Scanning instruments, such as
Landsat, may have comparable effects resulting from the cross-track scanning (Crippen, 1989).
Further major steps of radiometric preprocessing are atmospheric correction and correction of
topographically-induced illumination variations. These corrections are covered in section 4.3.3
under ‘Calculation of reflectance’. A number of techniques however try to avoid atmospheric
and topographic illumination corrections or unwanted atmospheric and topographic effects
through the algorithms applied, such as band ratios, normalized differences, or normalizations
within image matching that avoid or reduce such effects. It has been shown that in particular in
mountains with their complex topographic and atmospheric conditions, errors in according
corrections can be substantial and do not necessary improve uncorrected data as a whole (Paul,
2001).
Multi-angular or multitemporal applications may also be affected by the variation of the
bidirectional reflectance distribution (Figure 4.1). Such effects are difficult to correct, but can in
some cases be avoided by comparing data from similar acquisition positions and under similar
illumination conditions (see section 4.7.2).
5
Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
In addition to the above radiometric preprocessing, the construction of false color composites is
useful, in particular for manual glacier delineations or for manually correcting automatic
delineations (see 4.3.3).
Figure 4.1
Left: section of an ASTER 3N image over Svalbard, near Longyearbyen. Right:
Normalized difference index image between the orthorectified 3N and 3B. Green colours
indicate no difference between both images, red and blue indicate differences in digital
numbers. Differences are due to reflectance differences under different viewing directions,
i.e. effects from the bidirectional reflectance distribution function (BRDF) of the targets.
Certain landforms, e.g. glaciers and flood planes, can be especially well distinguished in the
difference image, indicating strong BRDF effects for these surface types in the NIR.
Topographic distortions and occlusions in the 3B image also contribute to the image
differences (e.g. steep north-facing flanks, shadows).
4.2.2 Geometric preprocessing
Here, we do not cover band-to-band or sensor-to-sensor co-registrations, which are usually done
at an early processing level of the satellite data. After this, the two most important types of
geometric pre-processing are the (i) transformation of the image data to a map projection, and (ii)
removal of topographic distortions. Depending on the processing level at which the base data are
provided, the image data may come in raw swath geometry, georeferenced (data in raw
geometry, but transformation information to map geometry given), or geocoded (transformed to
map geometry using a reference surface, usually the ellipsoid). Different data providers usually
have slightly different terminologies for their processing levels, or hybrid levels might exist (e.g.
ASTER L1B, which is georeferenced, and Earth rotation during overflight is already corrected
for).
6
Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
Removal of topographic distortions is usually done as orthoprojection (or orthorectification),
sometimes also as terrain correction applied to swath geometry data. This processing step
requires a DEM. The quality of the resulting orthoimages depends on
•
the quality of the spacecraft position and attitude information (i.e. of the pointing
direction);
•
the quality of ground control points (GCPs) in case such are used for defining the
image orientation, or for refining the latter as obtained from the sensor pointing data;
•
the quality of the DEM used.
It is important to keep in mind the error propagation from vertical DEM errors into horizontal
position errors in the resulting orthoimage. For a point with cross-track distance R from the nadir
track (orthogonal projection of the satellite path to the ellipsoid), an elevation error Δh causes for
a flying height above ground H a cross-track offset Δr :
∆r = ∆h
R
H
For scanners and pushbroom sensors the resulting offset is in cross-track direction. For frame
cameras, for instance, R and Δr are in radial direction from the image nadir point.
Whereas errors in map transformation can be roughly solved by co-registration of data to a
reference data set, errors in the topographic correction cannot be easily corrected due to the
usually non-systematic variations of DEM errors over a scene. Stereo sensors such as ASTER
offer the possibility to closely examine errors in orthoprojection (Kääb, 2004). Higher-order
geometric distortions may exist in the image data (Leprince et al., 2007; Nuth and Kääb, 2011),
but will for many applications not be corrected for.
For many glacier mapping applications, a final geolocation accuracy of the orthorectified data of
about one pixel will be sufficient, i.e. 30 m for the widely used Landsat TM/ETM data or
ASTER when including SWIR bands. Glacier changes to be observed have to exceed this
accuracy level roughly, which defines for a given rate in glacier-boundary change the time
interval at which those changes can be mapped at a statistically significant level.
For measuring glacier displacements using offset tracking between repeat image data, sometimes
a co-registration accuracy higher than one pixel is desirable, in particular for small movements
that require sub-pixel precision image-matching (section 4.8).
7
Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
4.3
Multispectral methods
4.3.1 Spectral response of typical glacial surfaces
The spectral signature of the ground cover determines the possibilities for their spectral
separability. Here, we focus on the VNIR and SWIR spectrum (Figure 4.2). Thermal infrared
(TIR) and the ground response of microwaves are briefly covered below in sections 4.4. and 4.5.
Schematic reflectance curves as a function of wavelength are given in Figure 4.3.
Snow
Fresh snow diffusely reflects up to 95% of the incoming VIS and about 50–80% of the NIR
radiation. In the VIS spectrum, snow reflectance decreases with particulate inclusions (e.g. dust,
soot, biota) (Warren and Wiscombe, 1980; Warren, 1982; Hall et al., 1989; Hall et al., 1990;
Painter et al., 2001; Takeuchi, 2009; Casey et al., 2012). In the near infrared the corresponding
influence of dust contamination decreases, and increasing snow grain size reduces snow
reflectance (Painter, 2011). In the shortwave infrared, the reflectance of snow is very low, with a
marked dependency on grain size, but a low one on snow contamination (Dozier, 1989;
Bourdelles and Fily, 1993). This large contrast between the VIS and SWIR spectral signature is
exploited for snow classification (e.g. Rott, 1976; Dozier, 1989). Under else similar emissivity
factors, the TIR and passive microwave emission of snow and ice is, in comparison to other
materials, limited by the fact that the surface temperature is at or below 0o C.
Ice
As a consequence of the reflectance properties of snow, bare glacier ice has a lower reflectance
than snow in the VIS spectrum due to the accumulation of optical active contaminants (Warren
and Brandt, 2008). This effect increases towards dirty glacier ice (Qunzhu et al., 1983;
Koelemeijer et al., 1993). In the NIR and SWIR, the dependency of reflectance on increasing
grain sizes from snow towards ice is effective. In addition, the presence of liquid water on the ice
surface might lead to reduced reflectivity in the NIR (Rott, 1976; Paul, 2004). For debris-covered
ice, the spectral signature of debris may prevail over the ice signature depending on the
percentage of debris-covered surface area within a pixel (Casey et al., 2012). If a glacier pixel is
covered with more than about 80 percents of a pixel by debris it cannot be separated from
periglacial debris pixels or surrounding bedrock by spectral methods alone.
Rock
Unlike snow and ice, rock or debris-covered surfaces show a significantly different reflectance in
the VNIR, SWIR and TIR. This fact usually allows for good spectral separability of such
surfaces against clean snow and ice. The highly variable reflectance of different mineral and rock
types in the VNIR, SWIR and TIR forms the base for lithology mapping from multi- and
hyperspectral imagery (e.g. Clark, 1999; Sabins, 1999; Rowan and Mars, 2003; Volesky et al.,
2003; Casey et al., 2012). Knowledge of the geology of a high mountain area supports
geomorphological and geomorphodynamical analyses in various ways (e.g. erosion processes)
(Casey et al., 2012). There are also direct applications for glaciological investigations. Detecting
the spatial distribution and the type of surface debris on glaciers, for instance, allows the
8
Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
identification of material origin and transport paths (Kääb, 2005b; Casey et al., 2012), and
therefore conclusions on present and past dynamics can be drawn.
Water
Reflection of open water in the VIS is highly variable depending on its turbidity. In the NIR and
SWIR, water strongly absorbs radiation, independent of its turbidity. NIR and SWIR reflection
of water is similar to snow and ice which can lead to misclassifications. Inclusion of the VIS is,
then, able to separate both categories (Huggel et al., 2002; Paul, 2004). Turbidity and
temperature of glacial lakes, important information for their characterization, can be derived
from VIS and thermal infrared data (Wessels et al., 2002).
Vegetation
Much research on multi- and hyperspectral remote sensing of vegetation is available (Liang,
2004; Jensen, 2007). Vegetation mapping usually takes advantage of high reflectivity in the NIR.
In glaciological research, the existence of vegetation itself might be the most interesting result
since it potentially points to, for instance, comparably stable surfaces, plant succession or lack of
surface abrasion. First-order vegetation mapping may be applied for excluding distinct areas
from further analyses or removing misclassifications, although in some rare cases, vegetation is
known to grow in soils overlying stagnant glacier ice.
Figure 4.2
Atmospheric transmission, sections of the optical and microwave spectrum, and
spectral range of Landsat ETM+, ASTER, and active microwave sensor bands. UV: ultraviolet;
VIS: visible; NIR: near infrared; SWIR: short-wave infrared; MIR: middle infrared; TIR:
thermal infrared, radar bands.
9
Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
Figure 4.3
Atmospheric transmission, locations of ASTER and Landsat bands, and typical
reflectance curves for glacier surfaces (left) and materials found around glaciers (right).
4.3.2 Principles of classification approaches
The surface signature in the spectral domain is used to describe and distinguish surface types and
conditions. Such classifications may be characterized (Schowengerdt, 2007) by terms such as:
hard / soft classification, manual / supervised / unsupervised classification, parametric / nonparametric classification, spatial / spectral segmentation, pixel / subpixel classification, and
multispectral / hyperspectral classification.
Hard versus soft classification
The result of a hard classification assigns exactly one category to a terrain point, providing sharp
boundaries between classes. In soft classification only likelihood values are given for a pixel to
belong to certain classes. In glacial environments, both discrete (hard) and fuzzy (soft)
transitions exist, thus, suggesting application of both classification types. In nature, the
boundaries between, for instance, glacier and rock, or snow and bare soil are in general very
sharp, while its detectability in remote sensing depends on the spatial image resolution. The
classification goal is, then, to determine the boundary position as accurately as possible. On the
other hand, the transitions, for instance, between stagnant ice and moraine, or gradually
decreasing vegetation coverage are smooth, so that a change in class likelihood might better
reflect the conditions in nature.
The choice of hard versus soft classification strategies depends not only on the natural
characteristics of the boundary between two surface categories, but also on the spectral
properties and spatial resolution of the sensor applied, and the scale considered. A category
transition in nature might be discrete in one part of the spectrum, but fuzzy in another part. If the
spatial resolution of a sensor (ground-projected instantaneous field of view, GIFOV) is
significantly larger than the spatial variations of a category, a resulting pixel contains a mixed
signal of more than one ground category (mixed pixel), independent of whether the category
transitions are discrete or fuzzy. Such mixed pixels can be classified hard (e.g. threshold) or soft
(e.g category percentages).
10
Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
Manual, supervised and unsupervised classification
Manual delineation of categories is done by combination of human interpretation and digitizing.
Manual delineation is used for simple classifications, but also for accuracy assessments, ground
truth acquisition, or completion or correction of other classifications. Supervised classification is
an application-driven method. Training areas of the categories to be mapped are operatorselected in order to develop the spectral signatures of these classes. Separability analysis is
applied to test if the selected categories can be distinguished at a statistically significant level
from another. Overlapping signatures are separated. The spectral signatures are then used to
automatically segment the entire image. During unsupervised classification the image is
automatically segmented in concentrations of spectral vectors, with the number of classes
assigned selected by the spectral analyst. These clusters of data classes represent artificial
categories. Unsupervised classification is, therefore, a data-driven method. The automatic
clusters must then be assigned to classes of user-interest. All three classification-strategies may
be combined.
Parametric versus non-parametric classification
Parametric classification assumes a certain statistical distribution of a particular class (e.g. mean
spectral vector and covariance). The statistical class parameters are estimated from training areas
within the image under analysis or inferred from other sources. An individual pixel is assigned to
a certain class according to its statistical proximity to the class parameters (e.g. nearest mean, or
maximum-likelihood). For non-parametric classification, the class-membership is decided not
on statistical parameters but on simple boundaries (e.g. boxes) in the spectral space defined
around training data, or on Euclidean distance to training pixels. Artificial neural networks
(ANN) are non-parametric classifications where the decision boundaries between the classes are
determined iteratively by minimizing an error criterion on the given training data (Jensen, 2005).
Spatial and spectral segmentation
Supervised and unsupervised, parametric and non-parametric classifications primarily rely on the
spectral characteristics of individual pixels. Image segmentation can also be done using
neighborhood relations of pixels (spatial segmentation). Edge detection or feature extraction
algorithms can be used to detect class boundaries. Areas of the same class membership may be
aggregated by region growing algorithms where individual pixels are joint according to algebraic
rules (e.g. based on spectral characteristics). Some spatial algorithms work also for panchromatic
imagery. The spectral and spatial approaches can be combined to spatial-spectral segmentation,
for instance, by applying spectral classification to pixel areas aggregated beforehand, by spatially
filtering the classification results, or – more advanced – by integrated object-oriented image
classifications (Jensen, 2005; Schowengerdt, 2007)
Subpixel classification
Hard classification methods assign each pixel to exactly one category. Quantifying the categories
contributing to the spectral signature of a mixed pixel leads to sub-pixel classification. With
subpixel accuracy a distinction must be made between spectral sub-pixel resolution (which
categories contribute, and how much?) and geometric sub-pixel resolution (where are the
categories?). In principle, it can not be resolved from a mixed pixel where the contributing
categories are located. Corresponding assumptions might, however, be drawn including the
11
Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
spatial context of a pixel or external knowledge, for instance, about the typical characteristics of
a boundary. Some kind of such geometric sub-pixel precision may be achieved by postclassification editing of class boundaries that were originally determined with pixel-precision.
Simple approaches of that type consist of interpolation or smoothing of classification results. For
instance, the pixel-wise (i.e. horizontally stepped) classification between glacier ice and adjacent
bedrock might be better represented by interpolating a smooth horizontal curve separating both
classes. In practice, however, such procedure is often complicated by a number of classification
problems, and might require knowledge-based interpolation algorithms; success of the approach
depends on the ratio between GIFOV and spatial resolution at which the category boundaries
should be mapped, among others.
Linear unmixing tries to determine for individual pixels the fraction of idealized pure signatures
(endmembers) contributing to its actual spectral composition. The endmember signatures can be
derived from extreme pixels assumed to consist only of one endmember class (i.e. pure unmixed)
or inferred from other data (Klein and Isacks, 1999; Painter et al., 2003). Fuzzy set classification
follows the opposite approach to the linear mixing concept by allowing one pixel to be member
of multiple categories with a certain probability connected to each membership (Binaghi et al.,
1997).
Combinations and others
The classification approaches listed here can be applied on multispectral data with a small
number of comparably broad bands as well as on hyperspectral data with a large number of
narrow bands. However, some of them are especially suited for hyperspectral data (e.g. linear
unmixing), and a number of additional algorithms exist for hyperspectral data (Lillesand and
Kieffer, 2000; Schowengerdt, 2007).
Classification approaches can be combined or applied sequentially. Instead of using only spectral
data, some classification algorithms also allow inclusion of spectral derivatives (e.g. band ratios
instead of bands themselves, or multitemporal data) or non-spectral data (e.g. DEM) (Brown et
al., 1998).
4.3.3 Image processing techniques for glaciers
Here, we summarize selected classification procedures in the optical spectrum that have already
been tested for glaciological applications, in particular mountain glaciers (cf. Sidjak and Wheate,
1999; Paul, 2001; Albert, 2002; Paul et al., 2002).
Manual delineation
Manual delineation of panchromatic or multispectral image features might be useful for highly
complex classifications where expert-knowledge is needed, for instance, for separating
rockglaciers, dead ice, debris-covered ice and periglacial debris from each other. An analyst is
able to include experience, knowledge and complex logical rules in the decision process, also
relying on non-spectral, i.e. multidimensional data or knowledge. Manual delineation is often
needed to complement and correct automatic classifications. (e.g. Rott and Markl, 1989; Hall et
al., 1992; Williams et al., 1997; Paul, 2002; Andreassen et al., 2008), but especially for older
12
Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
panchromatic data like air photos, Hexagon or Corona reconnaissance satellite images (e.g.
Bolch et al., 2010b; Narama et al., 2010) and those data without SWIR band (e.g. Landsat MSS,
Ikonos, Quickbird, ALOS AVNIR, etc.)
False color composites
False color composites (FCC) may take advantage of differences between the spectral signatures
of the categories prevailing in the multispectral image (Figure 4.4). For instance, in a Landsat
ETM+ 543 RGB-composite (red: channel 5, green: channel 4, blue: channel 3) snow and ice are
clearly separated from clouds, debris, rock or vegetation due to the significant step in reflectance
for ice and snow between VNIR and SWIR compared to the other materials. FCCs can be used
for facilitating manual delineation or for automatic classifications. They work well for clean ice
and snow (e.g. Della Ventura et al., 1983; Williams et al., 1991). Instead of FCCs also IHStransforms or decorrelation stretches might be helpful.
Figure 4.4
Landsat TM color composites over Tordrillo Mountains, Alaska. Left: TM
channels 3-2-1 red-gree-blue true color composite; middle: 432 false color composite; right:
543 false color composite.
Calculation of reflectance
Derivation of the ground reflectance at each pixel requires three steps (Bishop et al., 2004):
(1), calculation of the effective planetary reflectance at sensor (ρ) from the raw DNs of the
image. Calculation of planetary reflectance normalizes the at-sensor radiance (L sensor) for
exoathmospheric solar irradiance (L top-of-atmosphere) and depends thus for a given sensor on the
solar zenith angle and the Earth-Sun distance at acquisition time (ρ= L sensor / L top-of-atmosphere).
The DNs have to be transformed into radiance at the sensor using the calibration coefficients
gain and offset (Markham and Barker, 1985).
(2), atmospheric correction of the reflectance due to aerosol content (Vermote et al., 1997),
which improves the correspondence of satellite with in situ measurements.
(3), topographic correction to account for varying illuminations caused by different solar incident
angles, slope and aspect effects (Hugli and Frei, 1983; Sandmeier and Itten, 1997).
13
Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
Comparing the remote-sensing derived reflectance to that obtained from in situ measurements, or
to theoretical predictions from spectral signatures, allows to some extent for surface
interpretation from the corrected image. This approach has been used mostly to characterize
snow and ice facies, or their albedo, respectively (Hall et al., 1988; Gratton et al., 1990; Hall et
al., 1990; Gratton et al., 1993; Koelemeijer et al., 1993; Winther, 1993; Knap et al., 1999; Paul,
2001, 2004; Casey et al., 2012).
Spectral transforms
Intensity-hue-saturation transformation (IHS):
For visual interpretation and further
classification or processing, a transformation of RGB images (i.e. three-band imagery) into
another color space might be useful (Figure 4.5). The most often applied color space in that
context is the IHS color space. The IHS components of a RGB image are, for instance, obtained
from geometric projections of a color vector in the RGB space (Schowengerdt, 2007). The IHS
components of an image might be manipulated (e.g. replacement of intensity component by a
higher resolution I component, or stretching) and then inversely transformed to the RGB space.
Also, the components I, H or S might be used in a classification individually (Paul, 2004).
Principle component transformation (PC): Multispectral image bands are often highly correlated
due to the spectral similarity of the material within an entire wavelength range, due to
topographic effects (e.g. shading) and due to the spectral resolution of the applied sensor.
Principle component transformation (PCT) aims at transforming the original image linearly to
minimise the inter-band correlation (Richards, 1993; Schowengerdt, 2007):
By reducing the spectral image redundancy, PCs are able to facilitate classification from visual
inspection or other automatic approaches (Sidjak and Wheate, 1999). Direct assignment of PCTresults to classes of glaciological interest is, however, difficult (Paul, 2004) (Figure 4.6). PCs can
be useful for ice displacement measurements based on offset tracking in order to optimally
exploit all intensity variations contained in a multispectral data set (Ahn and Howat,
2011)(section 4.8). A disadvantage of this method is that the result varies with the spectral
classes present in an image.
Decorrelation stretching is an application of the PC or IHS transforms to reduce correlated and,
thus, redundant information of multispectral imagery (Gillespie et al., 1986; Schowengerdt,
2007). It allows for better visual exploitation of multispectral imagery and, thus, aids manual
delineation, and preparation and selection of further classification methods (Figure 4.9). For
decorrelation stretching the multispectral bands are PC transformed. The respective PCs are then
stretched to optimally fill the color-space, and transformed into the RGB color space (Figure
4.7). Decorrelation stretches are, for instance, available as ASTER Level 2 products.
14
Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
Figure 4.5
IHS transform of the Landsat TM 321 composite of Figure 4.4. Left: intensity
(brightness), middle: hue (color), right: saturation (amount of white).
Figure 4.6
The first three principle components (PC1, PC2, PC3, from left to right) of the
Landsat scene of Figure 4.4. All VIS, NIR and SWIR bands included in the transform.
15
Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
Figure 4.7
Decorrelation stretch of the Landsat scene of Figure 4.6, computed as the
histogram stretch of the PC1-PC2-PC3 red-green-blue composite.
Image algebra and segmentation
Different algebraic algorithms may be applied on spectral bands, among which band ratios R
Rij =
Rij =
DNi
, or
DN j
(DNi − DNmin(i) )
,
(DN j − DNmin( j ) )
and normalized differences indices (NDI, or modulation ratios) are the most widespread
(Schowengerdt, 2007)
NDIij =
Rij − 1 (DNi − DN j )
,
=
Rij + 1 (DNi + DN j )
where
DNi and DNj are digital numbers of two bands i and j that show high separability for the
category to be classified with respect to other categories in the image, and
DNmin are the minimum (i.e. darkest) DNs of an individual band.
For band ratios, subtraction of the minimum DN for each band applied reduces the illumination
effects from atmospheric scattering (Crippen, 1988). Similarly, the results of normalized
differences might be improved by subtraction of the band-specific minimum DNs. For ice and
snow, usually bands in the VNIR and bands in the SWIR are used in the index (e.g. TM 2,3 or 4
versus TM 5 Bayr et al., 1994; Rott, 1994; Jacobs et al., 1997; Paul, 2001, 2002; Paul et al.,
2002; , or ASTER 3 versus ASTER 4 Kääb et al., 2003a; Paul, 2004; Bolch and Kamp, 2006).
The normalized difference snow index (NDSI) is defined as (green–SWIR)/(green+SWIR)
(Dozier, 1989; Hall et al., 1995a; Sidjak and Wheate, 1999). Water can be detected by the
normalized difference water index (NDWI: (NIR–blue)/(NIR+blue), Huggel, 1998; Kääb et al.,
2000a; Kääb et al., 2000b; Huggel et al., 2002)), similar to the well-established NDVI for
vegetation (e.g. (NIR-red)/(NIR+red), Hardy and Burgan, 1999). Classification of water or
vegetation might be useful for eliminating misclassifications from the above band ratios for ice
and snow (Paul et al., 2002; Bolch and Kamp, 2006). Band ratios and NDIs partly eliminate
atmospheric and topographic influences that affect both bands applied in a similar way (e.g.
Holben and Justice, 1981). For final segmentation or classification of the ratio or normalized
difference image, thresholds have to be chosen. Image algebra and segmentation algorithms are
especially robust (Paul, 2001). To gain better control on the results further terms might be added,
possibly including additional bands.
16
Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
Currently VIS/SWIR or NIR/SWIR band ratios are among the most often and most operationally
used methods for mapping clean ice areas. A threshold on a visible channel is typically added to
improve the algorithm performance in cast shadow. The ratio VIS/SWIR misclassifies dark
shadow as glacier ice when using images with uncorrected path radiance. This can, however, be
removed with a threshold in the blue band (e.g. TM1). When little glacier ice is present in the a
scene the NIR/SWIR ratio is preferable (Paul and Kääb, 2005). A typical, ratio-based glacier
mask might therefore be:
DNVNIR
>k
DN SWIR
AND
DNVIS > l
where
DNVNIR are the digital numbers in visible or near-infrared channels, usually red or nearinfrared (e.g. TM3 or TM4, or AST2 or AST3; the choice of red or near-infrared depends on
the scene conditions);
DNSWIR are the digital numbers in a short-wave infrared channel (e.g. TM5 or AST4);
k is the threshold applied to the ratio in order to obtain a binary glacier mask; typical values
for k are on the order of 1.5-2.5, but may be even smaller or bigger and depend much on the
individual scene and imaging conditions;
l is the threshold on a short-wave visible channel (DNVIS ; e.g. TM1), with values that have to
be adjusted to the actual scene and illumination conditions (Paul and Andreassen, 2009).
Instead of DNVIS the intensity value of a IHS transform can be used (Paul, 2004).
Unsupervised classification
Unsupervised classifications tend to be successful for relatively homogeneous terrain with few
categories (e.g. clean-ice glaciers), whereas problems occur often for variable terrain with many
classes (e.g. clean ice, dirty ice, debris-covered ice, or clean ice in cast shadow, Paul, 2001,
2004). The unsupervised categories have to be assigned to classes of user-interest, which leaves
the major classification problems for terrain with low separability (e.g. the above ice types) still
to the analyst. Unsupervised ISODATA classification was applied by Aniya (1996) for glacier
inventorying of the Southern Patagonia Icefield using Landsat TM bands 1, 4 and 5.
Supervised classification
Supervised classification (e.g. maximum-likelihood, spectral angle mapper) works often
accurately for high mountain terrain with reasonable spectral separability. As for all spectral
classifications, debris-covered ice mapping remains challenging. Best results with supervised
classification can be achieved by including VIS, SWIR and TIR spectral bands as well as
derivatives such as band ratios, principle components or normalized differences, or TIR derived
silica weight percent (e.g. Bronge and Bronge, 1999; Sidjak and Wheate, 1999; Casey et al.,
2012). The choice of training areas is especially crucial for inhomogeneous terrain as often found
in alpine environments. Usually, the spectral class signatures trained from one scene cannot be
applied for other scenes, which reduces the automation capability or requires radiometric
adjustments. Further studies on supervised classification of mountain glaciers have been
performed by (Gratton et al., 1990; Binaghi et al., 1993; Paul, 2001, 2004). Klein and Isacks
17
Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
(1999) used spectral unmixing for classification of snow and ice, and Binaghi et al. (1997) fuzzy
set theory. Brown et al. (1998) applied maximum-likelihood and ANN classification on a DEM
of a glacierised landscape.
Artificial Neural Networks (ANN)
In ANN classification the decision boundaries are not fixed in a deterministic way from training
signatures, but obtained in an iterative fashion by minimizing an error criterion on the labelling
of the training data (Schowengerdt, 2007). ANN classification is not restricted to spectral data.
Other input data, such as spatial relations of pixels, multitemporal data, DEMs or DEM
derivatives can be used. For glacial and paraglacial terrain, ANN classifications from data of a
single domain (spectral or DEM) have been tested (Brown et al., 1998; Bishop et al., 1999; Paul
et al., 2004). The main potential of ANN application in glaciology might, however, lie in the
integration of multidimensional data (Paul et al., 2004).
Combinations
Often, classification procedures can (or must) be combined either by fusing different approaches
or by performing them sequentially. As mentioned for the supervised classification, spectral
derivatives or results of other pre-processing classifications may be combined as input layers.
Sidjak and Wheate (1999), for instance, achieved good results for glacier mapping from a
supervised maximum-likelihood classification using the PCs 2–4 of TM bands 1–7, a TM4/TM5
band ratio, and the NDSI. The most often used chain of sequential classification approaches is to
complete and correct automatic classifications manually. Such procedure is, for instance, so far
unavoidable for mapping debris-covered ice. Band ratios for glacier mapping, for instance, may
result in misclassifications for vegetation in shadow and turbid water, which in turn may be
eliminated by applying a NDVI and NDWI (Paul, 2004; Bolch and Kamp, 2006).
4.3.4 Post-processing and embedding in a GIS workflow
Spatial-domain filters may follow automatic segmentations and classifications. For instance, a
median filter on raw band ratio results is recommended (Paul et al., 2002). The filtered
segmentation results, still in raster format, are then converted to vector data. A number of
manual, semiautomatic or automatic steps has to follow in order to obtain glacier outlines from a
clean ice/snow mask (Kääb et al., 2002; Paul et al., 2002):
•
Separation of contiguous ice/snow areas into glaciers. For that step the preparation of
glacier basin masks was proposed (Paul et al., 2002). This step can be automated to a
high degree based on hydrological modeling using a DEM (Schiefer et al., 2008; Bolch et
al., 2010a).
•
Correction of the outlines for debris-covered ice, lakes (see above), and other
misclassifications
•
In case of repeat glacier inventorying, existing outlines can significantly facilitate the
latter step (Bolch et al., 2010a)
18
Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
•
Possibly, intersection of the glacier outlines with a DEM in order to derive topographic
glacier parameters (Paul et al., 2009).
•
Possibly, generation of central flow lines and intersection with the glacier outlines in
order to later derive glacier parameters related to this flow line (e.g. length) (Paul et al.,
2002)
4.4
Mapping of debris-covered ice
Mapping of debris-covered glacier parts constitutes one of the, if not the major bottleneck of
spectrally-based global-scale mapping of glacier areas. Before summarizing a number of
attempts to this problem, we would like to recall that, first, debris-covered glacier parts cannot be
sharply mapped in many cases even in principle, and that, second, debris-covered glacier parts
should be treated separately from clean-ice glacier parts in climate related glacier analysis, due to
the insulating properties of debris cover.
First, the transition from a debris-covered glacier part to debris-covered stagnant ice or to icecored moraines is ever-changing and often not sharp in nature. Possible definitions of this
transition might involve
•
ice content: A definition of the glacier boundary under debris might be deduced from the
percentage ice content at depth and its characteristics. The determination of intra- and
sub-debris ice is often the domain for in-situ geophysical surveys. Sub-surface ice content
is not strictly visible at the surface, but might have an effect accessible at surface such as
a thermal signal, distinct topography, or be detectable in repeat images;
•
ice ablation rates: The amount of ice ablation might define the glacier boundary
(Whalley, 1979; Haeberli and Epifani, 1986; Kääb et al., 1997). This ablation rate is not
necessarily related to the ice content, e.g. it is zero for an ice-cored moraine under
permafrost conditions where the supraglacial debris thickness exceeds the local active
layer thickness; thermokarst features and related topography are one expression of these
ablation rates;
•
kinematics of the debris-mantled ice body and its dynamic connection to the main glacier.
A deformation threshold would have to be set, which also implies the measurement
precision of the method used.
The above three basic factors might either be directly observed, or only their, perhaps combined,
expression at the surface, e.g. through topography, debris lithology, or velocity as detected in
repeat imagery.
Second, debris-covered glacier parts should be treated separately from clean-ice glacier parts
because of their potentially very different reaction to atmospheric forcing (Scherler et al., 2011).
Very thin debris-cover (< a few cm; (Østrem, 1959; Fukjii, 1977; Nakawo and Young, 1982;
Mattson et al., 1993; Rana et al., 1997; Adhikary et al., 2000; Kayastha et al., 2000; Nicholson
19
Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
and Benn, 2006)) increases ablation rates, whereas thicker debris-cover reduces ablation rates.
Where, however, glacier surface roughness increases and supraglacial ponds develop as a
consequence of the differential melt pattern and thermokarst processes induced by the debriscover, the area-averaged ablation rates might well be again on the order of clean ice or perhaps
even higher (Sakai et al., 2000; Bolch et al., 2011).
So far, attempts to map debris-covered glacier parts are based on topographic information,
spectral information, combinations of both, and radar interferometry (Shukla et al., 2010).
Due to glacial transport, supraglacial debris often has sources different from the rock and
sediments that are directly adjacent to a given position of supraglacial debris. In some cases, the
spectrally characterized lithological variations may therefore exhibit differences between
glacially transported or transformed debris and rock/debris from adjacent valley flanks (Kääb,
2005b; Casey et al., 2012). In by far most cases, however, VNIR and SWIR based spectral
classification alone will not be sufficient to map debris-covered glacier sections. The idea that
the thermal signal of debris and debris on ice could be different led to attempts to exploit TIR
bands (see section 4.5).
A number of studies tried to map debris-covered ice from geomorphometric parameters and
analysis, e.g. looking at distinct curvature changes at the glacier margin, DEM-based flow-paths,
or combining a number of DEM-derived parameters (Bishop et al., 2001; Bolch and Kamp,
2006). Further studies combined above information types (VNIR/SWIR or TIR +
geomorphometric parameters) (Paul et al., 2004). Other studies combined both VNIR and SWIR,
TIR and geomorphometric parameters (Bolch et al., 2007; Shukla et al., 2010; Bhambri et al.,
2011).
Another approach to mapping supraglacial debris is to compare elevation models of different
times and view elevation change over time as an indicator for ice ablation or ice-dynamic
processes (Kääb, 2005b). As yet, this approach is often hindered by the lack of available DEMs
with sufficient resolution and accuracy.
Ice kinematics can be directly measured using image matching methods (see section below) or
radar interferometry, thus potentially mapping the landscape units belonging to a, dynamically
defined, glacier system (Kääb, 2005b, a; Bolch et al., 2008). In an extension of this idea, the loss
of radar phase coherence over time indicates tiny ground movements and may surprisingly clear
depict the dynamic boundaries of a glacier. This approach seems especially successful for Lband radar data with comparably robust phase coherence for the terrain surrounding the glaciers
(Atwood et al., 2010; Strozzi et al., 2010) (Figure 4.8).
All above approaches proved to give useful results for particular regions and base data, but show
also clear deficiencies when transferred to other regions, either for reasons of data availability or
because of different glacier characteristics.
Manual digitizing of debris-covered glacier parts, thus so far still the most often used methods
for delineating debris-covered parts of glaciers, combines, depending on the operator and the
available data, a number of the above characteristics of glacier debris-cover, e.g. spectral,
geometric and topographic ones.
20
Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
Figure 4.8
North Iceland. Left: ALOS PALSAR phase coherence between the data of a 3-year
interferometric pair. Light blue indicates low coherence, pink to yellow high coherence.
Right: SPOT quicklook. The sharp transitions from low to high coherence coincide at many
places with the outlines of glaciers, debris-covered glaciers and rockglaciers.
4.5
Thermal imaging
So far the thermal infrared spectrum is little exploited for investigating mountain glaciers (Figure
4.8). Under incoming direct shortwave radiation, debris generally has stronger thermal emission
than ice due to its higher temperatures, leading to a strong signal in the TIR bands (Warren and
Brandt, 2008). It is presently under investigation, to what extent TIR data can be used to detect
glacier ice under loose or thin debris-cover from respective cooling of the superimposed debris
(Taschner and Ranzi, 2002; Kääb, 2005b; Mihalcea et al., 2008; Shukla et al., 2010; Casey et al.,
2012), (cf. Lougeay, 1974, 1982). Figure 4.9 shows an example on how the inclusion of TIR in
image interpretation and classification supports mapping of the geological composition of the
surface. Taschner and Ranzi (2002) show that differences in thermal infrared emission exist
between a debris-covered glacier and its surrounding, and how these differences can facilitate
glacier outline delineation. Shukla et al. (2010) and Bhambri et al. (2011) combine TIR data with
other spectral data and geomorphometric information for mapping debris-covered ice. In general,
mapping debris-covered ice from thermal data has the shortcoming that the recorded thermal
emission on such a surface is not strictly dependent on the ice underneath, but rather on a number
21
Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
of factors contributing to the energy balance such as roughness, incoming shortwave radiation
(see following paragraph), thermals emissivity of the surface layer, meteorological conditions,
etc. Kääb (2005b), and in more detail Casey et al. (2012), explore lithological variations on
supraglacial debris through TIR data, and how these results can be exploited glaciologically. TIR
based silica percent mapping may be used to indicate weathering and active turnover of glacier
debris (Casey et al., 2012). One disadvantage of TIR methods is TIR’s typically coarse spatial
resolution compared to VNIR and SWIR bands (ASTER: 90m, Landsat ETM: 60m).
During daytime, the longwave radiation emitted from the terrain surface is to a large extent
dependant on the shortwave incoming radiation (e.g. Mittaz et al., 2000; Hoelzle et al., 2001).
Therefore, nocturnal or early morning TIR imagery might potentially better indicate the thermal
differences of the ground material (Figure 4.8).
TIR data has also been used to map thermal resistance of debris cover (Suzuki et al., 2007).
In large-scale applications, low-resolution TIR data with high revisit times (in particular
MODIS) are operationally used for mapping the melt extent and its temporal variations over
large ice bodies (e.g. Greenland) (e.g. Hall et al., 2006).
Figure 4.9
Thermal infrared images over Tordrillo Mountains, Alaska, during the same day.
Left: daytime TIR from Landsat ETM, right: nighttime TIR from ASTER. During day (left),
longwave emission is modulated by topography due to incoming shortwave radiation. This
effect is largely reduced in the nighttime image (right) in which elevation-dependency of
longwave emission is more pronounced in addition to the emission contrasts due to surface
type. The difference between both data sets is also related to the ground thermal inertia.
22
Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
Figure 4.10 NIR-SWIR-TIR (ASTER bands 4-9-10) RGB composite over Hispar glacier,
Karakoram. Lithological differences become well visible in this band combination, and allow
for interpretation of the dynamic history of the glacier and the sources of supraglacial
debris. For instance, bends in distinct longitudinal debris bands point to ongoing or past
surges.
4.6
Microwave/SAR methods
Active microwave radar methods, where a beam of electromagnetic radiation is transmitted at an
oblique angle to the ground and the returned beam is captured and analyzed, can be used for
glacier mapping as well. In the microwave spectrum, the ground response is a function of the
applied wavelength, polarisation, incidence angle, ground roughness, microwave penetration
(volume scattering) and the complex dielectric constant of the surface (Lillesand and Kieffer,
2000). The latter describes the reflectivity and conductivity of the terrain material. The
reflectivity depends among other influences on the surface roughness in relation to the applied
wavelength and volume scattering. While centimeter-scale roughness appears rough for K-band
(λ≈2cm) (i.e., large fraction of diffuse reflection), it appears smooth in the L-band (λ≈20 cm)
(i.e., large fraction of specular reflection). The material conductivity in the microwave spectrum
is mainly dependent on the liquid water content. A cold and dry snow pack may be invisible for
microwaves so that the main reflection happens at the underlying material, whereas the
penetration into wet firn or ice is very small and surface reflectivity is high (Marshall et al.,
1995; Kelley et al., 1997; Engeset, 1999; Nagler and Rott, 2000; Zahnen et al., 2003; Warren and
Brandt, 2008). Small mountain lakes often represent a relatively smooth surface, thus, leading to
specular reflection and a dark signature in the SAR image (Pietroniro and Leconte, 2000).
Polarisation of the emitted and received radar signal (H: horizontal, V: vertical) is also able to
add to the separability of different categories in the SAR image, because different geometric
23
Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
properties of scatterers on and within the snow, firn and ice pack may lead to either preservation,
90° change or diffusion (de-polarisation) of polarized incoming radar waves with respect to the
backscattered ones (Figure 4.11, Figure 4.12).
Figure 4.11 Color composite of HH (sent H, received H) and HV channels (sent H, received V)
of an ALOS PALSAR winter scene over Kronebreen, Ny Ålesund, Salbard. HH: yellow, HV:
blue. Color differences in the composite occur due to changes in the degree to which
surface or near surface backscattering mechanisms change the polarization of the incoming
radar signal. Dominant co-polarization (here: HH) points to corner reflection and ice lenses
or layers, dominant cross-polarization (here: HV) points to volume scattering, e.g. in the
homogenous ice. North to the top.
24
Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
Figure 4.12 Color composite of a fully polarized ALOS PALSAR winter scene over Ny Ålesund,
Svalbard (Kronebreen to the lower right). North approx. to the right.
25
Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
Figure 4.13 Envisat ASAR backscatter over Ny Ålesund/Kronebreen, Svalbard. Upper right: 13
June 2008, middle left: 29 February 2008, middle right: 26 September 2008. The February
data are taken after a cold and snowy period (lower panel), the June data after a period of
slightly positive temperature with little precipitation, and the September data after a period
of strong rainfall. Whereas glacier delineation is difficult on single dates due to the different
26
Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
ground conditions, glacier edges become better visible in the multitemporal false color
composite of all the scenes (upper left). Lower panel: blue bars: precipitation, red line: air
temperature in Ny Ålesund (data from met.no). North to the top.
The penetration of radar waves into the snow, firn and ice pack, though unwanted for a number
of applications, may also be exploited, e.g. for the detection and mapping of crevasses covered
by snow (Figure 4.13).
Radar interferometry (interferometric synthetic aperture radar, or InSAR) for extraction of DEMs
is not covered in this chapter, but differential radar interferometry (DInSAR) for measurement of
glacier flow is briefly discussed in section 4.8.
Optical and microwave data for glacier remote sensing are in a number of aspects
complementary, so that their combination might offer new insights and approaches. These topics
are treated in the final section.
Figure 4.14 ALOS PALSAR winter backscatter data over Kronebreen, Ny Ålesund, Svalbard.
Upper image (PALSAR scene, cf. Figure 4.10), lower left: section of the same scene, lower
right: summer ASTER VNIR data over the same section as the lower left panel. Even through
27
Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
snow cover, crevasse zones (visible in the summer ASTER data) become clearly detectable
as high backscatter in the radar images.
4.7
Spectral change detection; multitemporal data merging
4.7.1 Overview
We term merging of either spectral or spatial domain data of different acquisition times as
multitemporal data merging. Two important multitemporal merging methods are: measurement
of elevation changes (Chapter 5) and terrain displacements (section 4.8). Merging of these two
categories of multitemporal geometry data allows for understanding and modelling of glacier
dynamics. Vertical changes and horizontal displacements are quantitatively combined with the
kinematic boundary condition at the surface (e.g. Kääb et al., 1998; Gudmundsson and Bauder,
1999; Kääb and Funk, 1999; Kääb, 2001).
A basic multitemporal merging in the spectral domain consists in the overlay of repeated image
data, most often applied for digital change detection (Singh, 1989; Mouat et al., 1993; Lillesand
and Kieffer, 2000). Change detection techniques include:
•
postclassification comparison,
•
multitemporal classification,
•
multitemporal principal component analysis,
•
multitemporal false colour composites,
•
algebraic expressions,
•
change vector analysis, and
•
change axis analysis.
Although mainly designed for application in the spectral domain, some of the above strategies
can also be applied for other data types.
Common spectral (and also non-spectral) classification procedures may be applied separately on
data sets of different times, and the classification results compared thereafter (postclassification
comparison). For instance, glacier-covered areas can be detected by multispectral classification
from a satellite image of time 1, and again from a satellite image of time 2. By simple algebraic
expressions the areas of glacier change can be extracted and quantified.
Instead of comparing results after the classification, terrain change may also be detected by
merging the multitemporal data within one classification procedure by defining change classes
and non-change classes (multitemporal classification).
The individual bands (or layers) of multitemporal data may be combined into one new multilayer
data set (Fujimura and Kiyasu, 1999). Principal component analysis of this new data set may
28
Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
then enable to detect changed terrain parts as little correlated components of the multilayer data
set. Such an approach is of special use when change has to be extracted from a large number of
data (Marshall et al., 1995).
Multitemporal false colour composites (FCCs) represent a powerful tool for visualising change
between two or three images (Figure 4.13, Figure 4.15, Figure 4.17). Such FCCs might directly
form the basis for mapping, or serve as preparation and evaluation of other change detection
algorithms. The number of data acquisition dates included in a multitemporal FCC is restricted to
three unless PCA is used beforehand to extract the most dominant changes from the data stack.
Multitemporal FCCs may also be used to visualize and investigate results of other change
detection techniques. (See separate section 4.7.2)
Algebraic expressions such as subtraction, ratios or normalized difference indexes between two
or more multitemporal data sets or derivatives of them are often used for change detection
(Figure 4.16, Figure 4.17). Detecting elevation differences from repeated DEMs is a simple
example for temporal data subtraction. Spectral differences between repeated imagery indicate
terrain changes, but are also affected by different illumination and atmospheric conditions. Under
some circumstances, ratios of multitemporal imagery tend to normalise for some effects such as
cast shadow (cf. Crippen, 1988). Figure 4.16 shows an example for detecting glacier movement
from repeated ASTER imagery from a multitemporal band ratio. Thereby, change was detected
from the displacement of individual terrain features such as crevasses (spatial change). Change
may also be detected from the general change in spectral signature, for instance when an area
becomes ice-free due to glacier retreat (spectral change) (Paul et al., 2004).
The above change detection approaches may be applied on the raw image data, but also on
corrected (e.g. for illumination and atmosphere), or further transformed or processed products
(e.g. orthoimagery, filter products, PCs). Application of change detection algorithms can also be
restricted to certain areas masked by some previous classification (Lillesand and Kieffer, 2000).
For a given pixel of a multispectral image or multilayer data set, two or more (spectral) variables
can be plotted against each other for individual acquisition dates. A change vector connects the
resulting points in the chosen variable space. The magnitude and direction of this vector (or
vector cluster for a group of pixels) may be characteristic for a certain type of change, for
instance plant succession in glacier forefields, development of the snow pack or glacier retreat.
(Lillesand and Kieffer, 2000).
A two- (or more-) dimensional scatter plot of a spectral band at time 1 versus the same band at
time 2 approximates the plot diagonal in case of no changes between the two acquisition dates.
For significantly changed pixels the respective plot points lie apart from this diagonal, forming
typical clusters for individual change types. Change axis analysis defines a non-change axis (the
above diagonal, or a parallel or slightly rotated axis) and perpendicular change axes. Threshold
coordinates in the change-axes space are then used to classify individual changed, or nonchanged, respectively, pixels (Lillesand and Kieffer, 2000). The method is also sensitive to
changes that the detection does not aim at, such as changes in shadow.
For all change detection approaches, accurate uniform geometry for the repeated data sets is
crucial. Such techniques are, thus, usually applied using imagery of the same sensor and same
29
Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
sensor position (e.g. repeat tracks) or orthoimages. Location differences between the compared
data sets might influence the change detection procedures heavily (Figure 4.16). Similarly,
illumination differences due to different sun positions may create significant apparent changes
(Figure 4.15, Figure 4.17).
Figure 4.15 Deposits of the 20 September 2002 rock/ice-avalanche at Karmadon, North
Ossetian Caucasus. Change detection is done by a multitemporal RGB-composite. R: ASTER
band 3 of 22 July 2001; G and B: ASTER band 3 of 13 October 2002. Avalanche track and
deposits, as well as lakes dammed by the deposits become well visible. Red-colored changes
in northern slopes are due to different shadow/illumination conditions between the
acquisition dates.
Figure 4.16 Normalized difference index image (right panel) of two ASTER images over a
glacier in Bhutan (20 January and 20 November 2001; left panel). Largest differences (black
and white colors) occur at crevasses due to their movement from left to right (black-andwhite pattern) and at the retreating calving front of the glacier. Differences at the lateral
30
Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
moraines point to errors in the orthorectification, presumably from DEM errors (SRTM)
propagating into horizontal orthoprojection errors.
Figure 4.17 Left: section of a RGB normalized difference index (NDI) of two Landsat scenes
(22 Sep 1992 and 31 Oct 2009). R: NDI of bands 5, G: NDI of bands 4, B: NDI of bands 2.
Right: section of the 2009 scene. The different dates of the year (> 1 month difference) lead
to illumination differences that become visible as apparent change (cf. section 4.7.2).
Glacier retreat, e.g. becomes visible as red colors. The white ring around the lake in the
middle (lake Sabai) reflects the reduction in lake area from the lake outburst on 3 Sep 1998.
4.7.2 Image Change Evaluation by Subtraction of Multispectral Anniversary Pairs
(ICESMAP)
Many chapters in this book show the results of change detection and landcover characterization
using multispectral images. Image differencing is a general processing technique developed to
identify uncorrelated zones within multitemporal image sets (Singh, 1989; Jensen, 2005), that is,
where surface changes have occurred during the time between image pair acquisitions. This perpixel image algebraic method simply involves subtraction of one digital image from another to
produce a third difference (or change) image. This can be expressed as,
Error! Bookmark not defined. Dxkij = Dxkij(t2) - Dxkij(t1) + C
31
Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
where x = pixel value for sensor band k; i and j are image line and column values respectively, t2
refers to the later image; t1 refers to the earlier image; and C is a constant for rescaling to
unsigned 8-bit image values (although radiometric bit depths may vary dependent upon sensor
characteristics). For multispectral imagery, the image spectral bands, k, will vary by choice of the
user, but for our work with ASTER imagery k includes the three VNIR bands (band3-NIR,
band2-red, band1-green) composited into RGB color space. Adequate change detection by
image differencing can be extremely challenging to do quantitatively if the images were acquired
under different illumination or sensor view conditions. Widely disparate shadowing and
complex scattering phase functions can be the most difficult or impossible challenges to remedy
(Figure 4.17).
Here we provide background to one method, Image Change Evaluation by Subtraction of
Multispectral Anniversary Pairs (ICESMAP), which was developed for GLIMS by the
University of Arizona group (J.S. Kargel and G. Leonard) in order to circumvent the need for
complex radiometric and geometric transformations of images over rugged surfaces. For this
book, ICESMAP has been applied to glaciers in the Chugach Mountains, Alaska (chapters XX &
XX); Hoodoo Mountain, British Columbia (chapter XX); Mount Rainier, Cascades (chapter
XX); the Mount Everest, Himalaya area (chapter XX); and the Mount Cook area, New Zealand
(chapter XX). This section is the first detailed presentation of the methodology. However, the
first published application of the ICESMAP methodology was for an image pair acquired over
Central East Greenland (Kargel et al., 2012).
Alpine glaciers and adjacent terrains typically undergo satellite-discernible surface changes due
to weather events, discrete geomorphological events, and seasonal and interannual fluctuations.
Some of these changes integrate effects of climate changes, ice motion, phase state, and mass
movements on glacier surfaces. Notable changes seen in differenced anniversary image pairs
may reveal changes in glacier extent and texture/microrelief; surface patterns and area density of
crevasses, debris, meltwater ponds and streams; snow distribution, and grain size of snow, firn,
and ice; melting or freezing of wet snow; snow and rock avalanches; and changes in vegetation
types and density on debris covered glaciers, moraines, and beds of drained glacier lakes. All of
these material or topographic changes may potentially be resolved spatially, spectrally, and/or
radiometrically, which is the core premise of image change detection: that surface variations
result in changes in radiance intensities which are larger in magnitude with respect to radiance
changes resultant from other sources unrelated to changes in the surface (Ingram et al., 1981).
Other sources in this case refer to relative changes in image-pair radiometry due primarily to
changes with little relation to surface changes. These may include variable view geometry and
scattering phase function of the surface; different sensor gains, shifting solar position in the sky
related to time of year and day, and differing atmospheric attenuation conditions related to
humidity and temperature, optically thin condensation clouds, and dust (Jensen, 1983). The
minimization or removal of these extraneous effects is required to see and quantify actual surface
changes; the approach of ICESMAP is simply to minimize these effects rather than correct for
them, noting that in practice, it is very difficult and often impossible to make an accurate
correction for all these variables. The problem with corrective approaches is that usually not
enough is known, or uncertainties are too great, to allow quantitative correction and isolation of
effects of actual changes. The ICESMAP solution is simple control of non-surface changes so
32
Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
that there is little or nothing to correct. If the sun is in the same position in the sky and the
sensor view of the surface remains the same, then there is no need to correct differences in solar
angle and differential effects of shadowing and hillshading, differential pathlength of photons
through the atmosphere, differential sensor gains, scattering phase functions, and so on. With
these factors controlled, it is better than having to correct for them. The limitation, of course, is
that images must be acquired on or near anniversary dates (or, as we explain below, otherwise
have the sun in the same place in the sky) with the same sensor and with the same view of the
surface. The trick is proper selection of adequate image pairs, which commonly but do not
always exist for a given area.
In some cases, single band image differencing may be sufficient to indicate where surface
changes have occurred, for example using either visible or NIR bands to identify surface
brightness and vegetation changes respectively. In this method, threshold DN values within the
difference image histogram are selected by the user to highlight areas of significant change (e.g.
statistically selected ‘area-change’ values: ≥ 2σ or ≤ -2σ). Note in this example that areas of
change are represented within the tails of the histogram; whereas values clustering on or near the
mean of the difference image histogram represent areas of little to no change. Whilst slightly
more complex, the differencing of multispectral image pairs can lead to a richer portrayal of
user-identifiable change features. This is true since multispectral data can better capture the
unique compositional information contained within the variable surface spectral responses
detected by the different channels of the sensor. The resultant change image may subsequently
contain information related to changing surface material compositions and their distributions.
The resulting difference image then requires post-differencing classification or coding of the
surface changes by an expert familiar with the terrain. Compared to single-band differencing
analysis, often based on thresholding, it may also be more complicated to segment specific
features of change in multiband differencing, and it may be necessary to identify what
combination of image bands are responsible for indicating a particular surface change before
additional thresholds or band ratios are applied. The interpretation can be but is not always
intuitive; if one image or both images contain saturated bands (where the upper range of
recorded radiance is limited by gains sensitivity), then the colors represented in the change image
may look dramatic but in fact can become difficult or impossible to interpret. This potential
problem underscores the importance of applying images that have been acquired with optimal
GLIMS gain settings (i.e., gain settings set to maximize optimal image quality for the glaciers
and features of interest for the particular area of study), or ignoring or very carefully dealing with
areas of the difference image where image saturation is problematic.
Principle factors affecting image quality needed for image differencing include atmospheric
conditions, cloud cover, sensor gain settings, and sun-target-sensor viewing geometry. For some
purposes, such as multitemporal mapping of the calving front extent of a glacier, there is little
preprocessing one must do to an image series except to assure proper image coregistration. In
many other cases, such as the mapping of more subtle landcover units or seeking to identify a
wide range of changing landcover / land use phenomena between images, it becomes necessary
to correct for various parameters of the changing observing geometry within the image pair
including sensor-solar azimuth angle and solar zenith angle, scattering phase functions,
atmospheric absorptions, differing sensor gain settings, and inter-instrument calibrations. This
can be done, but is sometimes a complex process that may ultimately prove futile if the images
33
Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
were acquired under conditions that are too disparate from one another. Subsequently, isolation
of seasonal versus interannual, versus long-term trends can be difficult. Therefore, for purposes
of assessment of interannual and long-term glaciological trends, change assessment is ideally
done with images acquired by the same imaging system in sun-synchronous orbits, using the
same instrument settings (e.g. gain) and viewing geometry , and acquired on (or near)
anniversary dates, and with the same atmospheric conditions. For these conditions, the solarsensor azimuth and solar zenith angle parameters are nearly identical; consequently observing
geometry and scene illumination are the same. If atmospheric conditions also are similar and
ground features are unchanged, then the same spectral reflectance image will be recorded,
because illumination, viewing, absorption, and scattering functions are the same. Furthermore, if
the instrument drift is known and corrected, then if one image is subtracted from the other, a
black field will be obtained (zero difference). Normally, of course, this is not the case as
surfaces are changing, and we are interested in these actual surface changes. In short, nearanniversary date image pairs better insure that spurious signals and artifacts, potentially resultant
from differing diurnal and seasonal sun angle effects, and seasonal vegetation phenological
differences are minimized.
For the current generation of orbiting Earth observing sensors, exact anniversary dates are rarely
acquired, but images acquired within a few days of the first image acquisition anniversary are
fairly common for more frequently imaged areas of Earth. Importantly, image pairs should be
selected for the season and or annual date(s) that may optimally capture the changing phenomena
of interest to the user. For satellite glacier studies, this may include images acquired at the end of
the regional ablation season. Secondarily yet also importantly, seasonality is important for image
date selections. Winter images may be relatively darker, contain significant shadowing, and be
hindered by increased snowcover. The former two problems may in part be mitigated by
acquiring image data from sun-synchronous sensors which collect scene data during local midday times. In addition, there is wider tolerance for non-exact anniversary date image pairs
collected near the solstices compared to the equinoxes, this because the sun’s observed path
through the sky is at a minimum during the solstices, and maximum during the equinoxes. As a
result, an image pair differing by one week collected near the solstice will show far less
shadowing differences than a similar one-week pair acquired near an equinox. Another approach
may include selection of images acquired over an equivalent number of days before and after a
solstice, so that the sun again is at the same place in the sky (though seasonal changes may
become significant). Shadows and surface scattering are not identical but are similar, and with
the same look direction the total solar radiative path length is also similar.
The success of image change assessment by this method is keyed to accurate image pair
coregistration. Serious mis-registration can produce ghosting and false indications of change,
and even mis-registration by a few tenths of a pixel can produce false outlining of features,
typically revealed where material contacts are highlighted by a pair of lines of differing colors
representing erroneous indications of change. If the registration is very precise, the generally
unchanging features (over time spans of years to decades) such as hills, valleys, mountains
peaks, and other relief features and various material units that are visually prominent in the
source images, virtually disappear as muted grayscale features in the change image. As a quality
minimum, image coregistration should produce RMSEs of less than or equal to ½ pixel, a
precision possible with a variety of commercial image processing software programs, or
34
Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
otherwise with user-generated image processing code. Georeferencing may also be required if
one or both images are not yet georeferenced and if it is important to assign geospatial values to
input and output products.
The following steps summarize the generalized approach for applying ICESMAP. Different
users may find that they can modify or eliminate some steps, depending upon the image data
type and quality used, the software or code applied, and the desired final products.
Generalized ICESMAP processing steps (e.g. ASTER VNIR):
Selection of VNIR multispectral image pair / series. Ideal images are cloud and haze free, apply
identical sensor gains, and are acquired on the anniversary date. However, images a few days off
the anniversary date usually provide useful results. Appendix 25.1 provides an error analysis for
one example of image differencing applied to an ASTER image pair acquired near Mount
Everest. ASTER Level 1B image data are well suited, with DN values in terms of scaled (8-bit)
radiance (Abrams et al., 1999).
1.
2.
3.
4.
5.
6.
Optional: Atmospheric correction or dark object subtraction from all bands (Chavez, 1975)
Produce 3-band image stacks (e.g. AST3-AST2-AST1 → RGB)
Subset images if necessary (i.e. reduce to area of interest)
Co-register image pair / series (< ½ pixel RMSExy)
Subtract earlier image from the later image
Re-scale to unsigned image format (full dynamic range of original imagery, e.g. 8-bit, 16bit). Note: this step may occur automatically depending on the software applied, or design of
the algorithm.
7. Render difference image into RGB space
8. Post-differencing, user-assignment of change features. User defines what all anomalous
change features represent.
9. Optional: Threshold individual or multiple difference image band(s) to segment and quantify
desired change features or apply a higher-level customized classifier or a clustering algorithm
such as Fuzzy C Means.
The change image from a pair of differenced 3-band images is also rendered as an RGB image
although the DN values are rescaled so that the lowest negative value is set to zero, and the
maximum change value is rescaled to DN = 255 (for 8-bit images). In this case, unchanging
features become a neutral gray, and changed surfaces become brighter or darker or colored.
However, there are usually portions of the difference image that show residual features that
cannot be completely removed, such as small variations on the surface where shadows have
shifted slightly, where differential slopes have slight illumination differences (especially when
sloping near grazing solar incidence). Changes in vegetation or soil or snow moisture can be of
interest, but differences in leaf moisture or slight differences in soil moisture and other fairly
insignificant changes can have large effects on a difference image. Typically though the change
features above are rather minor, against which more dramatic surface changes will stand out
boldly, particularly if the chosen image pair is of high quality. Real changes will be highlighted
in image brightness or color against a background of non-changing surfaces that appear
relatively subdued in the change image.
35
Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
In the chapters mentioned above, examples of difference image results and their interpretations
are given. The Himalayan image pair used for Chapter 25 has an extended explanation in terms
of method for that specific pair, interpretation, and some sources of error and artifacts. Other
chapters provide briefer explanations. In general, and not considering areas where one or another
image is saturated in one or more bands, areas where the differenced image is blue represent a
surface that has become bluer (or less red), such as formation of a pond; where reddish, they
have become more reddish (or less blue), such as drainage of a pond; and where greenish, band 2
has become brighter relative to bands 1 and 3. Where the image is black, it has become darker in
all bands; and white areas have become brighter in all bands. Recall one of the chief caveats that
when one or both of the images are saturated in one or more bands, the colors represented in the
change image are more difficult to interpret because some information is lacking in those pixels.
This most commonly occurs within areas that are highly reflective across all VNIR bands, such
as within or proximal to snowfields in the accumulation zones. Although such areas may appear
in a multitude of different hues, it is often still possible to identify the transient snowline (e.g.
ELA) especially if only one of the images in the image pair is less saturated or unsaturated. Table
X shows 13 typical cases of changing surface features within glaciated terrains, from the
perspective of how these features would appear in the before image and subsequently in the
difference image.
Table 4.1
Examples of ICESMAP image difference characteristics
36
Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
4.8
Ice flow
The kinematic boundary condition at the surface (Kääb and Funk, 1999) or the mass
conservation relation (Cogley et al., 2011) expresses that a change in terrain elevation over time
at a certain location and the vertical component of a three-dimensional displacement of an
individual particle at the surface describe different kinematic quantities. A change in elevation at
a defined glacier location may be the effect of mass balance, three-dimensional straining and/or
mass advection, the latter of which is again a function of glacier displacement and geometry.
Glacier displacements can be determined by different methods, though most of them only deliver
single components of the three-dimensional velocity vector:
•
vertical differences between multitemporal DEMs (see Chapter 5),
•
qualitative analysis of movements, such as from terrestrial indicators (e.g. lack of
vegetation; degree of lichen coverage and lichen size; surface weathering), change
37
Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
detection techniques (see above) or repeat image animation (flickering) (entire topic
not covered here) (Kääb et al., 2003b; Paul et al., 2007)
•
digital matching of repeated optical imagery (see below),
•
differential radar interferometry (DInSAR) (see below),
•
repeated terrestrial and satellite geodesy, i.e. repeat survey of natural or artificial
markers (not covered here),
•
analogue and analytical photogrammetric methods (not covered here),
•
DEM matching, i.e. digital matching of repeat DEM grids, similar to image matching
(not explicitly covered here, but implicitly under image matching)
•
terrestrial real or synthetic aperture radar (not covered here), and
•
mechanical methods such as strain wires (not covered here).
4.8.1 Image choice and pre-processing for image matching
A highly efficient method for measuring terrain displacements is the comparison of repeated
optical imagery. If original imagery is used, the obtained displacements have to be rectified using
the corresponding sensor model and orientation parameters (Kääb, 1996; Kääb et al., 1997; Kääb
and Funk, 1999; Kaufmann and Ladstädter, 2002). If orthoimages are used, the image
comparison directly delivers the horizontal components of the displacement vector. The
approach is, in principle, applicable to terrestrial, aerial, and spaceborne imagery.
The choice of images to be used for matching has to fulfill, or balance, the following conditions:
•
the time difference between the images used has to be large enough that the
displacements exceed the accuracy of the matches. The matching accuracy is mainly, but
not only (see section 4.8.4) a function of the image resolution and the precision of the
matching algorithm.
•
The time difference between the images has to be small enough that surface
transformations (melt, shearing, etc.) do not inhibit the finding of corresponding features
in the multitemporal images. Corresponding features may be destroyed after weeks for
fast-flowing glaciers under temperate climate conditions, or may last several years for
cold regions and/or slowly deforming debris-cover.
•
The larger the time difference between the images the better is the signal-to-noise ratio in
the estimated ice velocities.
•
The images have to be co-registered at an accuracy higher than the targeted measurement
accuracy or displacements. This co-registration can be part of the geometric preprocessing (e.g. orthorectification with common tie or ground control points) or a
38
Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
separate pre-processing step (i.e. co-registration using stable-ground points).
Alternatively, stable-ground offsets can be measured as part of the image matching
process, correction parameters estimated from these, and the displacements on glaciers
corrected accordingly. Note also remarks on orthoprojection in section 4.2.2, not least the
error propagation from DEMs into orthoimages.
Besides the geometric pre-processing of images to be matched (mainly co-registration),
radiometric pre-processing might be useful. In case the matching algorithm has no in-built
normalization between the gray values of the image, such processing will generally much
improve the results. Instead of a raster-based image matching process, points of interest with
sufficient gray-value variations around them can be detected and the matchings performed at
these locations only. Improved image matching results are reported if derivatives of the original
images are used, instead of the originals themselves, for instance gradient images (usually called
orientation images) or principal components (Frezzotti et al., 1998; Haug et al., 2010; Ahn and
Howat, 2011; Heid and Kääb, 2012a).
4.8.2 Image matching techniques
The digital comparison between multitemporal images, termed image matching, or intensity
matching/tracking or texture matching/tracking may be accomplished by area-based matching
techniques (also called block matching), or feature-based matching techniques. Block matching
compares complete grey-value arrays, i.e. image sections, to each other, feature-based matching
compares geometric forms such as edges or polygons that are extracted from the imagery and
converted to vector features. For measuring glacier movement, area-based methods seem
preferable due the fact that the reflectance variations to be tracked are often smooth, at least for
medium- and low-resolution data, and would thus not allow extraction of distinct vector features.
Algorithms for area-based matching operate in the spatial domain or in the frequency domain, or
a combination of both.
Among the spatial-domain techniques most suitable for glacier flow are double cross-correlation
and least-squares matching (Grün and Baltsavias, 1987; Scambos et al., 1992; Kääb and
Vollmer, 2000; Kaufmann and Ladstädter, 2002; Kääb, 2002; Debella-Gilo and Kääb, 2011a),
among the frequency-domain techniques are Fourier- and wavelet-based algorithms (Leprince et
al., 2007; Haug et al., 2010). All these area-based methods rely on extracting an image section
(reference template) from the reference image (e.g. time 1) and searching within a search
window the most similar search template in the search image (time 2). For frequency-domain
methods, the search window is equivalent to the search template.
A detailed list, explanation and comparison of the most commonly used matching methods in
glaciology is given by (Heid and Kääb, 2012a).
The size of the search window has to be chosen according to the expected maximum
displacement, such that the search template that corresponds with the reference template can, in
fact, be found in the search window. The size of the reference- and search templates have to be
chosen according to the textural characteristics of the ground surface. If the reference template
size is too small, the similarity surface has no clear maximum; if the reference template size is
39
Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
too large, computing time soars up, and deformations within the template area degrade the
matching accuracy or in extreme cases cause mismatches.
Some matching algorithms are inherently of sub-pixel precision such as least-squares or Fouriertransform based ones. For other algorithms such as cross-correlation, sub-pixel accuracy can be
achieved by either interpolating the images or templates to higher resolution (oversampling), or
by peak fitting to the similarity surface at its maximum (Debella-Gilo and Kääb, 2011a). In cases
where the observed displacement greatly exceeds the spatial resolution, sub-pixel precision
might not be necessary.
Digital motion measurements from repeated optical imagery have been applied
•
for ice sheets using satellite imagery (e.g. Scambos et al., 1992; Whillans and Tseng, 1995;
Frezzotti et al., 1998; Haug et al., 2010; Ahn and Howat, 2011),
•
for arctic glaciers using satellite imagery (e.g. Lefauconnier et al., 1994; Rolstad et al., 1997;
Dowdeswell and Benham, 2003; Kääb et al., 2006) (Figure 4.18),
•
for mountain glaciers using satellite, aerial or terrestrial imagery (e.g., Seko et al., 1998;
Nakawo et al., 1999; Evans, 2000; Kääb and Vollmer, 2001; Kääb, 2002; Kääb et al., 2003a;
Skvarca et al., 2003; Kääb, 2005a, b; Bolch et al., 2008; Scherler et al., 2008; Copland et al.,
2009; Quincey et al., 2009a; Quincey and Glasser, 2009; Heid and Kääb, 2012a, b).
Matching of orthoimages combined with repeated DEMs provides the horizontal displacements
and surface elevation changes. The vertical component of the displacement can be estimated
from the DEM gradients if surface-parallel displacement is assumed (Kääb and Funk, 1999). The
accuracy of this approximation is determined by the representativeness of the DEM used
compared to the actual terrain topography.
A combination of simultaneous multitemporal and multiangle image matching where a surface
particle is measured in all overlapping images of all image acquisition dates is, in fact, able to
deliver the three-dimensional surface particle displacements. Such procedure can be substantially
simplified by introducing approximated orthoimages (i.e., orthoimages computed from a coarse
DEM; also called pseudo-orthoimages) instead of the original imagery (Kaufmann and
Ladstädter, 2002; Leprince et al., 2007). The latter procedure and the above combined
orthoimage/DEM-gradient approach converge if DEMs with sufficiently high spatial resolution
are available. Repeat ASTER stereo-imagery, i.e. stereoscopic channels 3N and 3B, open up for
the possibility of simultaneous matching in both multitemporal and multiangular data for direct
measurement of three-dimensional displacements.
4.8.3 Post-processing and analysis
Any image matching results will contain inaccurate measurements or mismatches. If the
probability of such errors is not much reduced by pre-processing steps such as point-of-interest
40
Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
operators (Förstner, 2000; Kaufmann and Ladstädter, 2002; Debella-Gilo and Kääb, 2011b), a
number of post-processing steps are available to detect and remove erroneous measurements.
•
Minimum and maximum speeds (i.e. velocity magnitudes) to be accepted can be set.
•
Most matching algorithms provide also a measure of the matching quality, such as the
cross-correlation coefficients or signal-to-noise ratios. A threshold can be set to eliminate
results below a certain quality measure.
•
For single glaciers a directional slice (range of direction angles) can be chosen for
velocity vectors to be accepted. Similarly, a maximum directional offset set can be
chosen between the original displacement vectors and those from other sources, for
instance the aspect from a DEM.
•
Simple filters can be used to identify potentially erroneous measurements such as a
running vector median or RMS windows.
•
The original measurements can be compared to a low-pass filtered version of the
displacement field and a threshold set on the respective difference for measurements to
be accepted (Heid and Kääb, 2012a). A similar procedure can be built in into the
matching process itself to limit matching positions to most probable positions (Evans,
2000).
•
The original image 1 can be compared to a simulated image 1S that is obtained by
interpolating image 2 using the reverse displacements matched between 1 and 2. The
correlation between 1 and 1S is an indicator of the matching success, though only a
relative one, because the overall correlation level between 1 and 1S depends not only on
the displacement between times 1 and 2, but also different imaging conditions etc. The
method is thus particularly useful for algorithm comparisons (Debella-Gilo and Kääb,
2011b).
•
Instead of applying image matching to a single pair of raw or preprocessed images, a
number of versions of the input images can be matched (PCs, directional filtered, etc.)
and from the stack of results the most probably correct match be selected (Ahn and
Howat, 2011)
Glacier velocity fields from image matching can be analyzed in various ways, such as transverse
and longitudinal profiles, spatial gradients (i.e. ice deformation), temporal variations, stream line
and relative age calculations, direct inclusion in numerical flow models, etc. (Kääb, 2005b).
4.8.4 Accuracy
The total error budget of displacement vectors matched from repeat images contains
•
the pure precision of the matching algorithm, which may be on the order of up to 10-20%
of a pixel;
41
Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
•
the representativeness of a match for the matching window area covered: The measured
displacements are often assigned to the center pixel of the window, but can in theory
originate from anywhere inside the window, depending on the visual contrast and the
algorithm used. If the window contains deformation, this also introduces an error
(Debella-Gilo and Kääb, 2011b, 2012).
•
the geolocation accuracy of individual pixels, which might be on the same order of
magnitude as the pixel size itself, and stems from instabilities within the sensor (e.g.
mechanical scanner; sometimes called the co-registration accuracy) or instabilities of the
satellite that are not sufficiently captured by attitude measurements (e.g. jitter, Nuth and
Kääb, 2011; Heid and Kääb, 2012a);
•
the image co-registration or orthoprojection errors (e.g. from GCP or DEM errors);
•
the degree to which the tracked intensity variations actually reflect the underlying glacier
movement (e.g. surface transformations may change the intensity variations independent
of the movement);
•
the mismatches where a intensity feature in the search image, that does not correspond to
the same feature in the reference image, shows a higher similarity than the actually
correct corresponding features (e.g. self-similar crevasse or ogive patterns; snow dunes).
Obviously, where the images to be matched lack intensity variations that are sufficiently large
and stable over time, such as often the case for accumulation areas, image matching gets
unreliable or fails totally. As a result, the method will typically be suitable for ablation areas and
sections of accumulation areas with pronounced features such as crevasses, rock avalanche
deposits, etc. In addition, some matching methods are inherently more robust in low contrast
areas than others (Heid and Kääb, 2012a).
4.8.5 SAR offset tracking and interferometry
The phase differences in an interferogram constructed from two synthetic aperture radar (SAR)
images contains (i) the phase difference due to acquisition geometry (curved Earth phase;
ellipsoid), (ii) the phase difference due to actual topography above the ellipsoid (topographic
phase), (iii) the displacement phase due to any terrain shifts between the two acquisition times,
(iv) atmospheric phase distortions, and (v) noise (Figure 4.19). Removal of the curved Earth
phase from an interferogram, and subtraction of the topographic phase from the total phase
difference gives a differential interferogram with only displacement remaining apart from
atmospheric effects and noise (differential SAR interferometry, DInSAR). The topographic
phase can either be simulated from a DEM, or inferred from multiple interferograms the
difference between which is able to eliminate a constant displacement term (Bamler and Hartl,
1998).
Radar interferometry requires phase coherence between the acquisition times of the individual
SAR images. Such coherence over ice and snow is usually lost within hours or days, a fact that
limits the applicability of satellite radar interferometry usually to the 1-day tandem phase of ERS
42
Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
1 and ERS 2 during 1995 and 1996, and the occasional 3-day repeat cycle orbits of ERS (e.g.
1991), and a coordinated ERS-ENVISAT campaign. Longer periods of phase coherence of up to
one orbit cycle of typically several weeks or even longer are usually only found for ice bodies in
particularly stable cold and dry conditions (e.g. Canadian Arctic during winter, Antarctica, etc.).
New opportunities for radar interferometry over glaciers can be expected from the ESA Sentinel1 mission.
Radar interferometry provides only the line-of-sight component of the displacement. Other
components have to be projected, for instance assuming movement along the direction of
steepest descent as calculated from a DEM. Combination of interferograms from ascending and
descending orbits is thus able to provide two components of a three-dimensional displacement
vector, in particular for high latitudes with large azimuth differences between the two orbit
directions (Kwok and Fahnestock, 1996).
Numerous ice flow studies using radar interferometry exist, mainly in polar regions, but less
under typical mountain topography, for smaller glaciers, and in lower latitudes (e.g., Luckman et
al., 2007; Quincey et al., 2009b).
A totally different technique applicable to repeat SAR data is offset tracking (in principle
equivalent to above matching between optical images). Backscatter intensity or radar phase
texture is tracked between multitemporal SAR images using image matching techniques as
described above. As all image matching techniques, SAR offset tracking provides twodimensional displacements in the image plane. In case of phase coherence (see above) phase
texture can be tracked with high accuracy (radar speckle tracking). In case of coherence loss, i.e.
where radar interferometry is not anymore possible, there might still be sufficient SAR intensity
texture to be matched between the repeat images. The main differences between SAR offset
tracking and matching of repeat optical data consist (i) in the potential of SAR images to contain
backscatter intensity texture even in cases where optical images would not contain sufficient
texture (e.g. snow areas) and (ii) in the much higher noise level of SAR data (radar speckle). This
high noise level requires matching algorithms that are particularly robust against noise, and much
larger matching templates compared to optical data have to be employed. The latter requirement
limits the application of the method to larger glaciers. New high-resolution SAR sensors such as
Radarsat-2 or TerraSAR-X, however, allow glacier flow to be studied also for smaller glaciers by
reducing the geographic area of the matching templates (not the template size in number of
pixels), and by often showing more scene details to be potentially matched over time (Figure
4.20).
43
Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
Figure 4.18 Surface velocity field for a section of Kronebreen, derived from ASTER imagery of
26 June and 6 August 2001. Isolines indicate ice speed in meters per year. The surface
velocities of Kongsvegen are too small to be measured from repeated satellite imagery of
15 m resolution over a few weeks. Underlying ASTER image of 6 August 2001.
Figure 4.19 Radar interferogramme between an ERS 1/2 tandem pair of 5 and 6 Apr 1996.
Perpendicular baseline is 12 m, so that the topographic phase contribution is small and
44
Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
most visible color fringes due to ice movement. Kronebreen to the middle left shows
coherence loss due to the large velocities (cf. Figure 4.18). The large fringes to the middle
right indicate slowly deforming sea ice.
Figure 4.20 Surface displacements on Kronebreen derived from SAR offset tracking using
Radarsat2 finebeam data of 6 and 30 Apr 2009. Speeds increase from zero (lightblue/cyan)
over pink, yellow to blue (1.3 m/day).
4.9
Challenges, conclusions and perspectives
Classification approaches based on multispectral imagery are thoroughly developed and well
established. Significant progress can, however, be expected from newly developed combinations
of input layers and combinations of classification algorithms. For ice and snow applications,
inclusion of thermal bands is little tested and should be further investigated.
Paul (2002) and Paul (2004) found band ratios to be an especially simple, robust and fast method
for glacier mapping over large areas. Manual adjustment of the necessary threshold seems to be
preferable to thresholds automatically optimized from training areas (Rott and Markl, 1989). For
complicated situations adaptive threshold variations over one scene depending on variations of
the ground or illumination properties might be investigated.
There is still much potential in using spectral data and its derivatives alone, but multidimensional
classification including, for instance, DEMs may be most promising since they best reflect the
nature of high-mountain phenomena and processes. The increasing availability of suitable DEMs
supports this trend. The manual application of expert knowledge for delineation will be more and
more reduced but never become redundant (Rott and Markl, 1989).
Hyperspectral analysis of the glacial, peri- and paraglacial environment could provide new
advantages for classification, but also for understanding numerous phenomena and processes.
45
Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
While promising results are already available for geological applications, only little is known
about the hyperspectral response of alpine environments with rugged topography (Keller et al.,
1998; Schläpfer et al., 2000; Gruber et al., 2003; Casey et al., 2012). In contrast to multispectral
applications, hyperspectral ones have a large potential to lead much further beyond the human
ability for visual interpretation.
Microwave sensors respond especially sensitively to snow and ice and their physical properties
due to a varying liquid water content and varying penetration depths. However, much research is
still needed on the complex dielectric properties of snow and ice under various conditions to
allow for linking the recorded microwave response to geophysical characteristics of ground
materials.
Optical and microwave data for glacier remote sensing are in a number of aspects
complementary, so that their combination might offer new insights and approaches. In principle,
the combination of both spectral sections is possible within
•
DEM generation and elevation change detection: A major shortcoming of optical stereo
DEMs is their dependency on visual contrast. InSAR DEM generation is not affected by
that problem but rather relies on the (often sensitive) phase coherence. Optical DEMs can
serve as initial solution for an InSAR DEM by solving the modulo 2π ambiguity during
phase unwrapping, in particular over low-coherence areas. Such a DEM gets its overall
vertical scale from the auxiliary DEM and the details from radar interferometry.
Similarly, the absolute vertical scale of an InSAR DEM that is usually uncertain due to
baseline uncertainty can be refined by fitting the InSAR DEM to an auxiliary, e.g. optical
DEM or altimetry data (Moholdt and Kääb, 2012).
•
Quantification of glacier movement: Optimal conditions for differential radar
interferometry, radar offset tracking, and repeat optical image matching differ
significantly from each other, and may vary even over one scene at one point in time.
Radar interferometry requires phase coherence which is among others limited to small
displacements and deformations. Radar offset tracking does not suffer from such
problems, but is still more sensitive to ground changes than repeat optical data and
requires much larger matching windows. On the other hand, sufficient contrast of
backscatter features necessary for SAR offset tracking is usually found over much larger
areas than the visual contrast necessary for optical matching.
•
Spectral analysis of glaciers: Spectral fusion of optical and SAR data might be especially
promising for spectral analysis of glaciers (Rott and Strobl, 1992; Rott, 1994; Hall et al.,
1995b; König et al., 2001; Hall et al., 2009)(Warren and Brandt, 2008)
Acknowledgements
The contribution was supported by the ESA Climate Change Initiative project Glaciers_cci
(4000101778/10/I-AM), ESA GlobGlacier (21088/07/I-EC), and The Research Council of
Norway through the CORRIA project (No. 185906/V30).
46
Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
47
Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
References
Abrams, M., Hook, S. and Ramachandran, B. (1999): ASTER User Handbook, NASA document
prepared at JPL, California Institute of Technology.
Adhikary, S., Nakawo, M., Seko, K. and Shakya, B. (2000): Dust influence on the melting
process of glacier ice: experimental results from Lirung Glacier, Nepal Himalayas. In: M.
Nakawo, C.F. Raymond and A. Fountain (Eds.), Debris-covered glaciers. IAHS. 43-52.
Ahn, Y. and Howat, I.M. (2011): Efficient Automated Glacier Surface Velocity Measurement
From Repeat Images Using Multi-Image/Multichip and Null Exclusion Feature Tracking.
IEEE Transactions on Geoscience and Remote Sensing. 49(8), 2838-2846.
Albert, T. (2002): Evaluation of remote sensing techniques for ice-area classification applied to
the tropical Quelccaya ice cap, Peru. Polar Geography. 26(3), 210-226.
Andreassen, L.M., Paul, F., Kääb, A. and Hausberg, J.E. (2008): Landsat-derived glacier
inventory for Jotunheimen, Norway, and deduced glacier changes since the 1930s.
Cryosphere. 2(2), 131-145.
Aniya, M., Sato, H., Naruse, R., Skvarca, P. and Casassa, G. (1996): The use of satellite and
airborne imagery to inventory outlet glaciers of the Southern Patagonia Icefield, South
America. Photogrammetric Engineering and Remote Sensing. 62, 1361-1369.
Atwood, D.K., Meyer, F. and Arendt, A. (2010): Using L-band SAR coherence to delineate
glacier extent. Canadian Journal of Remote Sensing. 36, S186-S195.
Bamler, R. and Hartl, P. (1998): Synthetic aperture radar interferometry. Inverse Problems. 14,
R1-R54.
Bayr, K.J., Hall, D.K. and Kovalick, W.M. (1994): Observations on glaciers in the eastern
Austrian Alps using satellite data. International Journal of Remote Sensing. 15, 17331742.
Bhambri, R., Bolch, T. and Chaujar, R.K. (2011): Automated mapping of debris-covered glaciers
in the Garhwal Himalayas using ASTER DEMs and multi-spectral data. International
Journal of Remote Sensing.
Binaghi, E., Madella, A., Madella, P. and Rampini, A. (1993): Integration of remote sensing
images in a G. I. S. for the study of alpine glaciers. In: Winkler (Ed.), Remore Sensing
for Monitoring the Changing Environment of Europe. Balkema, Rotterdam. 173-178.
Binaghi, E., Madella, P., Montesano, M.P. and Rampini, A. (1997): Fuzzy contextual
classification of multisource remote sensing images. IEEE Transactions on Geoscience
and Remote Sensing. 35(2), 326-339.
Bishop, M., Bonk, R., Kamp Jr., U. and Shroder Jr., J.F. (2001): Terrain analysis and data
modeling for alpine glacier mapping. Polar Geography. 25(3), 182-201.
Bishop, M.P., Shroder, J.F. and Hickman, B.L. (1999): High resolution satellite imagery and
neuronal networks for information extraction in a complex mountain environment.
Geocarto International. 14, 17-26.
Bishop, M.P., Colby, J.D., Luvall, J.C., Quattrochi, D. and Rickman, D.L. (2004): Remotesensing science and technology for studying mountain environements. In: M.P. Bishop
and J.F. Shroder Jr (Eds.), Geographic Information Science and Mountain
Geomorphology. Springer, Praxis, Chichester, UK. 147-187.
48
Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
Bolch, T. and Kamp, U. (2006): Glacier mapping in high mountains using DEMs, Landsat and
ASTER data. Grazer Schriften der Geographie und Raumforschung. 41, 13-24.
Bolch, T., Buchroithner, M., Kunert, A. and Kamp, U. (2007): Automated delineation of debriscovered glaciers based on ASTER data. 27th EARSel Symposium. 403-410.
Bolch, T., Buchroithner, M.F., Peters, J., Baessler, M. and Bajracharya, S. (2008): Identification
of glacier motion and potentially dangerous glacial lakes in the Mt. Everest region/Nepal
using spaceborne imagery. Natural Hazards and Earth System Sciences. 8(6), 1329-1340.
Bolch, T., Menounos, B. and Wheate, R. (2010a): Landsat-based inventory of glaciers in western
Canada, 1985-2005. Remote Sensing of Environment. 114(1), 127-137.
Bolch, T., Yao, T., Kang, S., Buchroithner, M.F., Scherer, D., Maussion, F., Huintjes, E. and
Schneider, C. (2010b): A glacier inventory for the western Nyainqentanglha Range and
the Nam Co Basin, Tibet, and glacier changes 1976-2009. Cryosphere. 4(3), 419-433.
Bolch, T., Pieczonka, T. and Benn, D. (2011): Multi-decadal mass loss of glaciers in the Everest
area (Nepal Himalaya) derived from stereo imagery. The Cryosphere(5), 349-358.
Bourdelles, B. and Fily, M. (1993): Snow grain size determination from Landsat imagery.
Annals of Glaciology. 17, 87-92.
Bronge, L.B. and Bronge, C. (1999): Ice and snow-type classification in the Vestfold Hills, East
Antarctica, using Landsat-TM data and ground radiometer measurements. International
Journal of Remote Sensing. 20, 225-240.
Brown, D.G., Lusch, D.P. and Duda, K.A. (1998): Supervised classification of types of glaciated
landscapes using digital elevation data. Geomorphology. 21, 233-250.
Casey, K., Kääb, A. and Benn, D. (2012): Geochemical characterization of supraglacial debris
via in situ and optical remote sensing methods: a case study in Khumbu Himalaya, Nepal.
The Cryosphere. 6, 85-100.
Chavez, P.S.J. (1975): Atmospheric, solar, and MTF corrections for ERTS digital imagery.
American Society of Photogrammetry, Fall Technical Meeting. 69.
Clark, R.N. (1999): Spectroscopy of rocks and minerals, and principles of spectroscopy. In: A.N.
Rencz (Ed.), Remote Sensing for the Earth Sciences. Manual of Remote Sensing, 3rd
Edition. John Wiley and Sons, New York. 3-58.
Cogley, J.G., Hock, R., Rasmussen, L.A., Arendt, A., Bauder, A., Braithwaite, P., Jansson, P.,
Kaser, G., Möller, M., Nicholson, L. and Zemp, M. (2011): Glossary of glacier mass
balance and related terms. IHP-VII Technical Documents in Hydrology. 86. UNESCOIHP.
Copland, L., Pope, S., Bishop, M.P., Shroder, J.F., Clendon, P., Bush, A., Kamp, U., Seong, Y.B.
and Owen, L.A. (2009): Glacier velocities across the central Karakoram. Annals of
Glaciology. 50(52), 41-49.
Crippen, R.E. (1988): The dangers of underestimating the importance of data adjustments in
band rationing. International Journal of Remote Sensing. 9(4), 767-776.
Crippen, R.E. (1989): A Simple Spatial-Filtering Routine for the Cosmetic Removal of ScanLine Noise from Landsat Tm P-Tape Imagery. Photogrammetric Engineering and
Remote Sensing. 55(3), 327-331.
Debella-Gilo, M. and Kääb, A. (2011a): Sub-pixel precision image matching for measuring
surface displacements on mass movements using normalized cross-correlation. Remote
Sensing of Environment. 115(1), 130-142.
49
Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
Debella-Gilo, M. and Kääb, A. (2011b): Locally adaptive template sizes in matching repeat
images of Earth surface mass movements. ISPRS Journal of Photogrammetry and
Remote Sensing. 69, 10-28.
Debella-Gilo, M. and Kääb, A. (2012): Measurement of surface displacement and deformation of
mass movements using least squares matching of repeat high resolution satellite and
aerial images. . Remote Sensing. 4(1), 43-67.
Della Ventura, A., Rampini, A., Rabagliati, R. and Serandrei Barbero, R. (1983): Glacier
monitoring by satellite. Il Nuovo Cimento. C-1(6), 211-221.
Dowdeswell, J.A. and Benham, T.J. (2003): A surge of Perseibreen, Svalbard, examined using
aerial photography and ASTER high-resolution satellite imagery. Polar Research. 22(2),
373-383.
Dozier, J. (1989): Spectral signature of alpine snow cover from Landsat 5 TM. Remote Sensing
of Environment. 28, 9-22.
Engeset, R.V. (1999): Comparison of annual changes in winter ERS-1 SAR images and glacier
mass balance of Slakbreen, Svalbard. International Journal of Remote Sensing. 20(2),
259-271.
Evans, A.N. (2000): Glacier surface motion computation from digital image sequences. IEEE
Transactions on Geoscience and Remote Sensing. 38(2), 1064-1072.
Frezzotti, M., Capra, A. and Vittuari, L. (1998): Comparison between glacier ice velocities
inferred from GPS and sequential satellite images. Annals of Glaciology. 27, 54-60.
Fujimura, S. and Kiyasu, S. (1999): Classification of terrain objects using hyperdimensional
(multitemporal multispectral) images through purpose-oriented feature extraction. IEEE
Geoscience and Remote Sensing Symposium, IGARSS’99. 1192-1194.
Fukjii, Y. (1977): Field experiment on glacier ablation under a layer of debris cover. Journal of
the Japanese Society of Snow and Ice. 39, 20-21.
Förstner, W. (2000): Image preprocessing for feature extraction in digital intensity, color and
range images. In: A. Dermanis, A. Grün and F. Sansò (Eds.), Geomatic Methods for the
Analysis of Data in the Earth Sciences. Lecture Notes in Earth Sciences. Springer. 165189.
Gillespie, A.R., Kahle, A.B. and Walker, R.E. (1986): Color enhancement of highly correlated
images: I. Decorrelation and HSI contrast stretches. Remote Sensing of Environment. 20,
209-235.
Gratton, D.J., Howarth, P.J. and Marceau, D.J. (1990): Combining DEM parameters with
Landsat MSS and TM imagery in a GIS for mountain glacier characterization. IEEE
Transactions on Geoscience and Remote Sensing. 28(4), 766-769.
Gratton, D.J., Howarth, P.J. and Marceau, D.J. (1993): Using Landsat-5 Thematic Mapper and
digital elevation data to determine the net radiation field of a mountain glacier. Remote
Sensing of Environment. 43, 315-331.
Gruber, S., Schläpfer, D. and Hoelzle, M. (2003): Snow-free ground albedo derived from DAIS
7915 hyperspectral imagery for energy balance modeling in high-alpine topography. 8th
International Conference on Permafrost, Extented Abstracts. 47-48.
Grün, A. and Baltsavias, E.P. (1987): High-precision image matching for digital terrain
modelling. Photogrammetria. 42, 97-112.
Gudmundsson, G.H. and Bauder, A. (1999): Towards an indirect determination of the massbalance distribution of glaciers using the kinematic boundary condition. Geografiska
Annaler. 81A(4), 575-583.
50
Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
Haeberli, W. and Epifani, F. (1986): Mapping the distribution of buried glacier ice - an example
from Lago Delle Locce, Monte Rosa, Italian Alps. Annals of Glaciology. 8, 78-81.
Hall, D., Chang, A.T.C., Foster, J.L., Benson, C.S. and Kovalick, W.M. (1989): Comparison of
in situ and Landsat derived reflectances of Alaskan glaciers. Remote Sensing of
Environment. 28, 493-504.
Hall, D.K., Chang, A.T.C. and Siddalingaiah, H. (1988): Reflectances of glaciers as calculated
using Landsat 5 Thematic Mapper data. Remote Sensing of Environment. 25(3), 311-321.
Hall, D.K., Bindschadler, R.A., Foster, J.L., Chang, A.T.C. and Siddalingaiah, H. (1990):
Comparison of in situ and satellite derived reflectances of Forbindels Glacier, Greenland.
International Journal of Remote Sensing. 11, 493 - 504.
Hall, D.K., Williams, R.S. and Bayr, K.J. (1992): Glacier recession in Iceland and Austria. EOS
Transactions, American Geophysical Union. 73(12), 129,135,141.
Hall, D.K., Riggs, G.A. and Salomonson, V.V. (1995a): Development of methods for mapping
global snow cover using Moderate Resolution Imaging Spectroradiometer (MODIS) data.
Remote Sensing of Environment. 53, 127-140.
Hall, D.K., Williams, R.S. and Sigurdsson, O. (1995b): Glaciological observations on
Bruarjökull, Iceland, using synthetic aperture radar and Thematic Mapper satellite data.
Annals of Glaciology. 21, 271-276.
Hall, D.K., Williams, R.S., Casey, K.A., DiGirolamo, N.E. and Wan, Z. (2006): Satellitederived, melt-season surface temperature of the Greenland Ice Sheet (2000-2005) and its
relationship to mass balance. Geophysical Research Letters. 33(11), -.
Hall, D.K., Nghiem, S.V., Schaaf, C.B., DiGirolamo, N.E. and Neumann, G. (2009): Evaluation
of surface and near-surface melt characteristics on the Greenland ice sheet using MODIS
and QuikSCAT data. Journal of Geophysical Research-Earth Surface. 114, -.
Hardy, C.C. and Burgan, R.E. (1999): Evaluation of NDVI for monitoring moisture in three
vegetation types of the western U.S. Photogrammetric Engineering and Remote Sensing.
65(5), 603-610.
Haug, T., Kääb, A. and Skvarca, P. (2010): Monitoring ice shelf velocities from repeat MODIS
and Landsat data - a method study on the Larsen similar to C ice shelf, Antarctic
Peninsula, and 10 other ice shelves around Antarctica. Cryosphere. 4(2), 161-178.
Heid, T. and Kääb, A. (2012a): Evaluation of existing image matching methods for deriving
glacier surface displacements globally from optical satellite imagery. Remote Sensing
and Environment. 118, 339-355.
Heid, T. and Kääb, A. (2012b): Repeat optical satellite images reveal widespread and long term
decrease in land-terminating glacier speeds. The Cryosphere. 6, 467-478.
Hoelzle, M., Mittaz, C., Etzelmüller, B. and Haeberli, W. (2001): Surface energy fluxes and
distribution models of permafrost in European mountain areas: an overview of current
developments. Permafrost and Periglacial Processes. 12(1), 53-68.
Holben, B.N. and Justice, C.O. (1981): An examination of spectral band rationing to reduce the
topographic effect on remotely sensed data. International Journal of Remote Sensing.
2(2), 115-133.
Huggel, C. (1998): Periglaziale Seen im Luft- und Satellitenbild. Diploma Thesis, University of
Zurich.
Huggel, C., Kääb, A., Haeberli, W., Teysseire, P. and Paul, F. (2002): Satellite and aerial
imagery for analysing high mountain lake hazards. Canadian Geotechnical Journal. 39(2),
316-330.
51
Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
Hugli, H. and Frei, W. (1983): Understanding Anisotropic Reflectance in Mountainous Terrain.
Photogrammetric Engineering and Remote Sensing. 49(5), 671-683.
Ingram, K., Knapp, E. and Robinson, J.W. (1981): Change detection technique development for
improved urbanized area delineation. Technical memorandum CSCITM-81/6087,
Computer Sciences Corporation, Silver Springs, Maryland, U.S.A.
Jacobs, J.D., Simms, E.L. and Simms, A. (1997): Recession of the southern part of Barnes Ice
Cap, Baffin Island, Canada, between 1961 and 1993, determined from digital mapping of
Landsat TM. Journal of Glaciology. 43(143), 98-102.
Jensen, J.R. (1983): Urban/suburban land use analysis. Manual of Remote Sensing. 2. Virginia
American Society of Photogrammetry, Falls Church.
Jensen, J.R. (2005): Introductory digital image processing: a remote sensing perspective. 3rd
edition. Pearson Prentice Hall.
Jensen, J.R. (2007): Remote sensing of the environment: an earth resource perspective. 2nd
edition. Pearson Prentice Hall.
Kargel, J.S., Ahlstrøm, A.P., Alley, R.B., Bamber, J.L., Benham, T.J., Box, J.E., Chen, C.,
Christoffersen, P., Citterio, M., Cogley, J.G., Jiskoot, H., Leonard, G.J., Morin, P.,
Scambos, T., Sheldon, T. and Willis, I. (2012): Brief Communication: Greenland’s
shrinking ice cover: “fast times” but not that fast. The Cryosphere. 6, 533-537.
Kaufmann, V. and Ladstädter, R. (2002): Spatio-temporal analysis of the dynamic behaviour of
the Hochebenkar rock glaciers (Oetztal Alps, Austria) by means of digital
photogrammetric methods. Grazer Schriften der Geographie und Raumforschung. 37,
119-140.
Kayastha, R.B., Takeuchi, Y., Nakawo, M. and Ageta, Y. (2000): Practical prediction of ice
melting beneath various thickness of debris cover on Khumbu Glacier, Nepal, using a
positive degree-day factor. In: M. Nakawo, C.F. Raymond and A. Fountain (Eds.),
Debris-covered glaciers. IAHS. 71-81.
Keller, P., Keller, I. and Itten, K.I. (1998): Combined hyperspectral data analysis of two Alpine
lakes using CASI and DAIS 7915 imagery. EARSeL workshop. CD-ROM.
Kelley, R., Engeset, R., Kennett, M., Barrett, E.C. and Theakstone, W. (1997): Characteristic
snow and ice properties of a Norwegian ice cap determined from complex ERS SAR. 3rd
ERS Symposium (ESA). CD-ROM.
Klein, A.G. and Isacks, B.L. (1999): Spectral mixture analysis of Landsat thematic mapper
images applied to the detection of the transient snowline on tropical Andean glaciers.
Global and Planetary Change. 22(1-4), 139-154.
Knap, W.H., Brock, B.W., Oerlemans, J. and Willis, I.C. (1999): Comparison of Landsat-TM
derived and ground-based albedos of Haut Glacier d'Arolla, Switzerland. International
Journal of Remote Sensing. 20(17), 3293-3310.
Koelemeijer, R., Oerlemanns, J. and Tjemkes, S. (1993): Surface reflectance of Hintereisferner,
Austria, from Landsat 5 TM imagery. Annals of Glaciology. 17, 17-22.
Kwok, R. and Fahnestock, M.A. (1996): Ice sheet motion and topography from radar
interferometry. IEEE Transactions on Geoscience and Remote Sensing. 34(1), 189-200.
Kääb, A. (1996): Photogrammetrische Analyse zur Früherkennung gletscher- und
permafrostbedingter Naturgefahren im Hochgebirge. Mitteilungen der Versuchsanstalt
für Wasserbau, Hydrologie und Glaziologie der ETH Zürich. 145, Zürich.
52
Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
Kääb, A., Haeberli, W. and Gudmundsson, G.H. (1997): Analysing the creep of mountain
permafrost using high precision aerial photogrammetry: 25 years of monitoring Gruben
rock glacier, Swiss Alps. Permafrost and Periglacial Processes. 8(4), 409-426.
Kääb, A., Gudmundsson, G.H. and Hoelzle, M. (1998): Surface deformation of creeping
mountain permafrost. Photogrammetric investigations on rock glacier Murtèl, Swiss
Alps. 7th International Permafrost Conference. 531-537.
Kääb, A. and Funk, M. (1999): Modelling mass balance using photogrammetric and geophysical
data. A pilot study at Gries glacier, Swiss Alps. Journal of Glaciology. 45(151), 575-583.
Kääb, A., Haeberli, W., Huggel, C. and Paul, F. (2000a): Glacier- and permafrost-related hazards
in high mountains: integrative assessment in the Swiss Alps based on remote sensing and
geo-information systems. X Congresso Peruano de Geologia. CD-ROM.
Kääb, A., Huggel, C. and Paul, F. (2000b): Früherkennung hochalpiner Naturgefahren mittels
Fernerkundung. Interpraevent 2000. 1. 49-60.
Kääb, A. and Vollmer, M. (2000): Surface geometry, thickness changes and flow fields on
creeping mountain permafrost: automatic extraction by digital image analysis. Permafrost
and Periglacial Processes. 11(4), 315-326.
Kääb, A. (2001): Photogrammetric reconstruction of glacier mass balance using a kinematic iceflow model: a 20-year time-series on Grubengletscher, Swiss Alps. Annals of Glaciology.
31, 45-52.
Kääb, A. and Vollmer, M. (2001): Digitale Photogrammetrie zur Deformationsanalyse von
Massenbewegungen im Hochgebirge. Vermessung Photogrammetrie Kulturtechnik.
99(8), 538-543.
Kääb, A. (2002): Monitoring high-mountain terrain deformation from air- and spaceborne optical
data: examples using digital aerial imagery and ASTER data. ISPRS Journal of
Photogrammetry and Remote Sensing. 57(1-2), 39-52.
Kääb, A., Paul, F., Maisch, M., Hoelzle, M. and Haeberli, W. (2002): The new remote sensing
derived Swiss glacier inventory: II. First results. Annals of Glaciology. 34, 362-366.
Kääb, A., Huggel, C., Paul, F., Wessels, R., Raup, B., Kieffer, H. and Kargel, J. (2003a): Glacier
monitoring from ASTER imagery: accuracy and applications. EARSel eProceedings. 2,
43-53.
Kääb, A., Isakowski, Y., Paul, F., Neumann, A. and Winter, R. (2003b): Glaziale und
periglaziale Prozesse: Von der statischen zur dynamischen Visualisierung.
Kartographische Nachrichten. 53(5), 206-212.
Kääb, A. (2004): Mountain glaciers and permafrost creep. Methodical research perspectives from
earth observation and geoinformatics technologies. Habilitation Thesis, University of
Zurich.
Kääb, A. (2005a): Combination of SRTM3 and repeat ASTER data for deriving alpine glacier
flow velocities in the Bhutan Himalaya. Remote Sensing of Environment. 94(4), 463474.
Kääb, A. (2005b): Remote Sensing of Mountain Glaciers and Permafrost Creep. Schriftenreihe
Physische Geographie. Glaziologie und Geomorphodynamik. 48. University of Zurich,
Zurich.
Kääb, A., Lefauconnier, B. and Melvold, K. (2006): Flow field of Kronebreen, Svalbard, using
repeated Landsat7 and ASTER data. Annals of Glaciology. 42, 7-13.
König, M., Winther, J.-G. and Isaksson, E. (2001): Measuring snow and ice properties from
satellite. Reviews of Geophysics. 39(1), 1-27.
53
Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
Lefauconnier, B., Hagen, J.O. and Rudant, J.P. (1994): Flow speed and calving rate of
Kronebreen glacier, Svalbard, using SPOT images. Polar Research. 13(1), 59-65.
Leprince, S., Ayoub, F., Klinger, Y. and Avouac, J.P. (2007): Co-Registration of Optically
Sensed Images and Correlation (COSI-Corr): an operational methodology for ground
deformation measurements. Igarss: 2007 Ieee International Geoscience and Remote
Sensing Symposium, Vols 1-12, 1943-1946.
Liang, J. (2004): Quantitative remote sensing of land surfaces. Wileys.
Lillesand, T.M. and Kieffer, R.W. (2000): Remote Sensing and image interpretation - fourth
edition. Wiley, New York etc.
Lougeay, R. (1974): Detection of buried glacial and ground ice with thermal infrared remote
sensing. In: H.S. Santeford and J.L. Smith (Eds.), Advanced Concepts and Techniques in
the Study of Snow and Ice Resources. National Academy of Sciences, Washington D.C.
487-494.
Lougeay, R. (1982): Landsat thermal imaging of alpine regions. Photogrammetric Engineering
and Remote Sensing. 48(2), 269-273.
Luckman, A., Quincey, D. and Bevan, S. (2007): The potential of satellite radar interferometry
and feature tracking for monitoring flow rates of Himalayan glaciers. Remote Sensing of
Environment. 111(2-3), 172-181.
Markham, B.L. and Barker, J.L. (1985): Spectral characterization of the Landsat Thematic
Mapper sensors. International Journal of Remote Sensing. 6(5), 697-716.
Marshall, G.J., Rees, W.G. and Dowdeswell, J.A. (1995): The discrimination of glacier facies
using multitemporal ERS - 1 SAR data. In: J. Askne (Ed.), Sensors and environmental
applications of remote sensing. Balkema, Rotterdam. 263-269.
Mattson, L., Gardner, J. and Young, G. (1993): Ablation on debris covered glaciers: an example
from the Rakhiot glacier, Punjab, Himalaya, Snow and Glacier Hydrology. IAHS. 289296.
Mihalcea, C., Mayer, C., Diolaiuti, G., D'Agata, C., Smiraglia, C., Lambrecht, A., Vuillermoz, E.
and Tartari, G. (2008): Spatial distribution of debris thickness and melting from remotesensing and meteorological data, at debris-covered Baltoro glacier, Karakoram, Pakistan.
Annals of Glaciology, Vol 48. 48, 49-57.
Mittaz, C., Hoelzle, M. and Haeberli, W. (2000): First results and interpretation of energy-flux
measurements of Alpine permafrost. Annals of Glaciology. 31, 275-280.
Moholdt, G. and Kääb, A. (2012): A new DEM of the Austfonna ice cap by combining
differential SAR interferometry with ICESat laser altimetry. Polar Research, in press.
Mouat, D.A., Mahin, G.G. and Lancaster, J. (1993): Remote sensing techniques in the analysis of
change detection. Geocarta International. 2(39-50).
Nagler, T. and Rott, H. (2000): Retrieval of wet snow by means of multitemporal SAR data. Ieee
Transactions on Geoscience and Remote Sensing. 38(2), 754-765.
Nakawo, M. and Young, G.J. (1982): Estimate of Glacier Ablation under a Debris Layer from
Surface-Temperature and Meteorological Variables. Journal of Glaciology. 28(98), 2934.
Nakawo, M., Yabuki, H. and Sakai, A. (1999): Characteristics of Khumbu Glacier, Nepal
Himalaya: recent change in the debris-covered area. Annals of Glaciology. 28, 118-122.
Narama, C., Kaab, A., Duishonakunov, M. and Abdrakhmatov, K. (2010): Spatial variability of
recent glacier area changes in the Tien Shan Mountains, Central Asia, using Corona
54
Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
(similar to 1970), Landsat (similar to 2000), and ALOS (similar to 2007) satellite data.
Global and Planetary Change. 71(1-2), 42-54.
Nicholson, L. and Benn, D.I. (2006): Calculating ice melt beneath a debris layer using
meteorological data. Journal of Glaciology. 52(178), 463-470.
Nuth, C. and Kääb, A. (2011): Co-registration and bias corrections of satellite elevation data sets
for quantifying glacier thickness change. The Cryosphere. 5, 271-290.
Painter, T.H., Duval, B., Thomas, W.H., Mendez, M., Heintzelman, S. and Dozier, J. (2001):
Detection and quantification of snow algae with an airborne imaging spectrometer.
Applied and Environmental Microbiology. 67(11), 5267-5272.
Painter, T.H., Dozier, J., Roberts, D.A., Davis, R.E. and Green, R.O. (2003): Retrieval of
subpixel snow-covered area and grain size from imaging spectrometer data. Remote
Sensing of Environment. 85(1), 64-77.
Painter, T.H. (2011): Comment on Singh and others, 'Hyperspectral analysis of snow reflectance
to understand the effects of contamination and grain size'. Journal of Glaciology. 57(201),
183-185.
Paul, F. (2001): Evaluation of different methods for glacier mapping using Landsat TM.
EARSeL eProceedings. 1, 239-245.
Paul, F. (2002): Changes in glacier area in Tyrol, Austria, between 1969 and 1992 derived from
Landsat 5 Thematic Mapper and Austrian Glacier Inventory data. International Journal of
Remote Sensing. 23(4), 787-799.
Paul, F., Kääb, A., Maisch, M., Kellenberger, T. and Haeberli, W. (2002): The new remotesensing-derived Swiss glacier inventory: I. Methods. Annals of Glaciology. 34, 355-361.
Paul, F. (2004): The new Swiss glacier inventory 2000 – Application of remote sensing and GIS.
PhD Thesis, University of Zurich.
Paul, F., Huggel, C. and Kääb, A. (2004): Mapping of debris-covered glaciers using
multispectral and DEM classification techniques. Remote Sensing of Environment. 89(4),
510-518.
Paul, F. and Kääb, A. (2005): Perspectives on the production of a glacier inventory from
multispectral satellite data in Arctic Canada: Cumberland Peninsula, Baffin Island.
Annals of Glaciology, Vol 42, 2005. 42, 59-66.
Paul, F., Kaab, A. and Haeberli, W. (2007): Recent glacier changes in the Alps observed by
satellite: Consequences for future monitoring strategies. Global and Planetary Change.
56(1-2), 111-122.
Paul, F. and Andreassen, L.M. (2009): A new glacier inventory for the Svartisen region, Norway,
from Landsat ETM plus data: challenges and change assessment. Journal of Glaciology.
55(192), 607-618.
Paul, F., Barry, R.G., Cogley, J.G., Frey, H., Haeberli, W., Ohmura, A., Ommanney, C.S.L.,
Raup, B., Rivera, A. and Zemp, M. (2009): Recommendations for the compilation of
glacier inventory data from digital sources. Annals of Glaciology. 50(53), 119-126.
Pietroniro, A. and Leconte, R. (2000): A review of Canadian remote sensing applications in
hydrology, 1995-1999. Hydrological Processes. 14, 1641-1666.
Quincey, D.J., Copland, L., Mayer, C., Bishop, M., Luckman, A. and Belo, M. (2009a): Ice
velocity and climate variations for Baltoro Glacier, Pakistan. Journal of Glaciology.
55(194), 1061-1071.
55
Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
Quincey, D.J. and Glasser, N.F. (2009): Morphological and ice-dynamical changes on the
Tasman Glacier, New Zealand, 1990-2007. Global and Planetary Change. 68(3), 185197.
Quincey, D.J., Luckman, A. and Benn, D. (2009b): Quantification of Everest region glacier
velocities between 1992 and 2002, using satellite radar interferometry and feature
tracking. Journal of Glaciology. 55(192), 596-606.
Qunzhu, Z., Meisheng, C., Xuezhi, F., Fengxian, L., Xianzhang, C. and Wenkun, S. (1983): A
study of spectral reflection characteristics for snow, ice and water in the north of China.
In: B. Goodison (Ed.), Hydrological applications of remote sensing and remote data
transmission. IAHS. 451-462.
Rana, B., Nakawo, M., Fukushima, Y. and Ageta, Y. (1997): Application of a conceptual
precipitation-runoff model (HYCYMODEL) in a debris Langtans Valley, Nepal
Himalaya. Annals of Glaciology. 25, 226-231.
Richards, J.A. (1993): Remote sensing digital image analysis - an introduction. Second edition.
Springer, Berlin.
Rolstad, C., Amlien, J., Hagen, J.O. and Lundén, B. (1997): Visible and near-infrared digital
images for determination of ice velocities and surface elevation during a surge on
Osbornebreen, a tidewater glacier in Svalbard. Annals of Glaciology. 24, 255-261.
Rott, H. (1976): Analyse der Schneeflächen auf Gletschern der Tiroler Zentralalpen aus LandsatBildern. Zeitschrift für Gletscherkunde und Glazialgeologie. 12(1), 1-18.
Rott, H. and Markl, G. (1989): Improved snow and glacier monitoring by the Landsat Thematic
Mapper. Workshop on Landsat Themaic Mapper applications. SP-1102. ESA. 3 - 12.
Rott, H. and Strobl, D. (1992): Synergism of SAR and Landsat TM imagery for thematic
mapping in complex terrain. Advances in Space Research. 12(7), 425-431.
Rott, H. (1994): Thematic studies in alpine areas by means of polarimetric SAR and optical
imagery. Advances in Space Research. 14(3), 217-226.
Rowan, L.C. and Mars, J.C. (2003): Lithologic mapping in the Mountain Pass, California area
using Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER)
data. Remote Sensing of Environment. 84(3), 350-366.
Sabins, F.F. (1999): Remote sensing for mineral exploration. Ore Geology Reviews. 14(3-4),
157-183.
Sakai, A., Takeuchi, N., Fujita, K. and Nakawo, M. (2000): Role of supraglacial ponds in the
ablation process of a debris-covered glacier in the Nepal Himalayas. In: M. Nakawo, C.F.
Raymond and A. Fountain (Eds.), Debris-covered glaciers. IAHS, Seattle. 119-130.
Sandmeier, S. and Itten, K. (1997): A physically-based model to correct atmoshperic and
illumination effects in optical satellite data of rugged terrain. IEEE Transactions on
Geoscience and Remote Sensing. 35(3), 708-717.
Scambos, T.A., Dutkiewicz, M.J., Wilson, J.C. and Bindschadler, R.A. (1992): Application of
image cross-correlation to the measurement of glacier velocity using satellite image data.
Remote Sensing of Environment. 42(3), 177-186.
Scherler, D., Leprince, S. and Strecker, M.R. (2008): Glacier-surface velocities in alpine terrain
from optical satellite imagery - Accuracy improvement and quality assessment. Remote
Sensing of Environment. 112(10), 3806-3819.
Scherler, D., Bookhagen, B. and Strecker, M.R. (2011): Spatially variable response of
Himalayan glaciers to climate change affected by debris cover. Nature Geoscience. 4,
156-159.
56
Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
Schiefer, E., Menounos, B. and Wheate, R. (2008): An inventory and morphometric analysis of
British Columbia glaciers, Canada. Journal of Glaciology. 54(186), 551-560.
Schläpfer, D., Bojinski, S., Schaepman, M. and Richter, R. (2000): Combination of geometric
and atmospheric correction for AVIRIS data in rugged terrain. AVIRIS workshop.
Schowengerdt, R.A. (2007): Remote Sensing. Models and Methods for Image Processing - Third
Edition. Academic Press, San Diego.
Seko, K., Yakubi, H., Nakawo, M., Sakai, A., Kadota, T. and Yamada, Y. (1998): Changing
surface features of Khumbu glacier, Nepal Himalayas revealed by SPOT images. Bulletin
of Glacier Research. 16, 33-41.
Shukla, A., Arora, M.K. and Gupta, R.P. (2010): Synergistic approach for mapping debriscovered glaciers using optical-thermal remote sensing data with inputs from
geomorphometric parameters. Remote Sensing of Environment. 114(7), 1378-1387.
Sidjak, R.W. and Wheate, R.D. (1999): Glacier mapping of the Illecillewaet icefield, British
Columbia, Canada, using, Landsat TM and digital elevation data. International Journal of
Remote Sensing. 20, 273-284.
Singh, A. (1989): Digital change detection techniques using remotely sensed-data. International
Journal of Remote Sensing. 10(6), 989-1003.
Skvarca, P., Raup, B. and De Angelis, H. (2003): Recent behaviour of Glaciar Upsala, a fastflowing calving glacier in Lago Argentino, southern Patagonia. Annals of Glaciology. 36,
184-188.
Strozzi, T., Paul, F. and Kääb, A. (2010): Glacier mapping with ALOS PALSAR Data within the
ESA GlobGlacier Project ESA Living Planet Symposium.
Suzuki, R., Fujita, K. and Ageta, Y. (2007): Spatial distribution of thermal properties on debriscovered glaciers in the Himalayas derived from ASTER data. Bulletin of Glaciological
Research 24, 13-22.
Takeuchi, N. (2009): Temporal and spatial variations in spectral reflectance and characteristics
of surface dust on Gulkana Glacier, Alaska Range. Journal of Glaciology. 55(192), 701709.
Taschner, S. and Ranzi, R. (2002): Comparing the opportunities of Landsat-TM and ASTER data
for monitoring a debris covered glacier in the Italian Alps within the GLIMS project.
IEEE Geoscience and Remote Sensing Symposium. 4. 1044-1046.
Vermote, E., Tanré, D., Deuze, J.L., Herman, M. and Morcette, J.J. (1997): Second simulation of
the satellite signal in the solar spectrum: an overview. IEEE Transactions on Geoscience
and Remote Sensing. 35(3), 675-686.
Volesky, J.C., Stern, R.J. and Johnson, P.R. (2003): Geological control of massive sulfide
mineralization in the Neoproterozoic Wadi Bidah shear zone, southwestern Saudi Arabia,
inferences from orbital remote sensing and field studies. Precambrian Research. 123(2),
235-247.
Warren, S.G. and Wiscombe, W.J. (1980): A Model for the Spectral Albedo of Snow .2. Snow
Containing Atmospheric Aerosols. Journal of the Atmospheric Sciences. 37(12), 27342745.
Warren, S.G. (1982): Optical properties of snow. Reviews of Geophysics and Space Physics. 20,
67-89.
Warren, S.G. and Brandt, R.E. (2008): Optical constants of ice from the ultraviolet to the
microwave: A revised compilation. Journal of Geophysical Research-Atmospheres.
113(D14).
57
Kääb et al. (2014): Glacier mapping and monitoring based on spectral data. Pages 75-112.
In: Kargel, Leonard, Bishop, Kääb, Raup (Eds.): Global Land Ice Measurements from Space. Springer Praxis.
Wessels, R., Kargel, J.S. and Kieffer, H.H. (2002): ASTER measurement of supraglacial lakes in
the Mount Everest region of the Himalaya. Annals of Glaciology. 34, 399-408.
Whalley, W.B. (1979): Relationship of Glacier Ice and Rock Glacier at Grubengletscher, Kanton
Wallis, Switzerland. Geografiska Annaler Series a-Physical Geography. 61(1-2), 49-61.
Whillans, I.M. and Tseng, Y.-H. (1995): Automatic tracking of crevasses on satellite images.
Cold Regions Science and Technology. 23(2), 201-214.
Williams, R.S., Hall, D.K. and Benson, C.S. (1991): Analysis of glacier facies using satellite
techniques. Journal of Glaciology. 37, 120-127.
Williams, R.S., Hall, D.K., Sigurdsson, O. and Chien, J.Y.L. (1997): Comparisson of satellitederived with ground-based measurements of the fluctuations of the margins of
Vatnajökull, Iceland, 1973-1992. Annals of Glaciology. 24, 72-80.
Winther, J.G. (1993): Landsat TM derived and in situ summer reflectance of glaciers in
Svalbard. Polar Reserach. 12, 37-55.
Zahnen, N., Jung-Rothenhäusler, F., Oerter, H., Wilhelms, F. and Miller, H. (2003): Correlation
between Antarctic dry snow properties and backscattering characteristics in RADARSAT
SAR imagery. EARSeL eProceedings. 2, 140-148.
Østrem, G. (1959): Ice melting under a thin layer of moraine, and the existence of ice cores in
moraine ridges. Geografiska Annaler Series a-Physical Geography. 41, 228-230.
58