DATA SET DOCUMENTATION
1. Title
1.1 Airborne Tracking Sunphotometer (ATSP) data
1.4 March 19, 1993
2. Investigator
2.1 Principal Investigator: Michael Spanner, NASA Ames Research Center,
Co-Investigators: Robert Wrigley, Rudolf Pueschel, Philip Russell and
John
Livingston, NASA Ames Research Center.
2.2 Measurement of aerosol optical properties and atmospheric-correction
of remotely sensed data acquired during Hapex-II/Sahel.
2.3 Contacts
2.3.1 Michael Spanner
Robert Wrigley
2.3.2 NASA Ames Research Center
Mail Stop 242-4
Moffett Field CA 94035 USA
NASA Ames Research Center
Mail Stop 242-4
Moffett Field CA 94035 USA
2.3.3 415-604-3620
415-604-6060
2.3.4 Spanner@gaia.arc.nasa.gov
Wrigley@gaia.arc.nasa.gov
2.3.5 415-604-4680
415-604-4680
John Livingston
NASA Ames Research Center
Mail Stop 245-5
Moffett Field CA 94035
415-604-3386
Livingston@gal.arc.nasa.gov
2.4 Please acknowledge the NASA Ames Research Center Investigation,
Michael Spanner, Principal Investigator if these data are used or
referenced
3. Introduction
3.1 The goal of this overall investigation was to measure aerosol optical
properties from both ground and aircraft based Sunphotometers during
the Hapex-II/Sahel experiment. These measurements are to be used to:
1) measure the magnitude and variability of the aerosol optical depth in
both time and space; 2) Determine the optical properties of the Sahelian
aerosols; and 3) atmospherically correct some remotely sensed data
acquired during the Hapex-II/Sahel experiment. The data described in
this documentation are the Airborne Tracking Sunphotometer data.
3.2 The phenomenon being measured is the atmospheric aerosol optical
depth. The parameters include Rayleigh optical depth, aerosol optical
depth, time, latitude, longitude, airmass, solar elevation and altitude.
3.3 We made measurements of aerosol optical depth with the ATSP
during 11 days of the Hapex-II/Sahel experiment. These days were
August 29, September 1, 2, 3, 6, 8, 9, 10, 12, 13, and 17. Variability
in
atmospheric optical properties across some of the flight lines stresses
the
need for an instrument that can measure this spatial variation. These
measurements were made at the same time that NS001 TMS and ASAS
data were acquired. The overall quality of the data appear to be
excellent. The instrument was calibrated after the experiment at Mauna
Loa Observatory, Hawaii.
The Airborne Tracking Sunphotometer data will be used in conjunction
with the Field Sunphotometer data to determine the magnitude and
variability of the aerosol optical depth in both time and space. From
the
aerosol optical depth, we will invert the data using an algorithm
developed by King et al., 1978, to derive the size distribution of the
Sahelian aerosols. We will then use Mie theory to calculate the aerosol
phase function and single scattering albedo. Finally, we will use the
atmospheric correction algorithm of Wrigley et al., 1992 to
atmospherically
correct selected NS001 TMS, ASAS and SPOT data collected during the
Hapex-II/Sahel experiment.
Atmospheric correction of SPOT data will utilize the aerosol properties
derived from surface optical depth measurements. Atmospheric correction
of NS001 and ASAS data will utilize aerosol properties derived from the
airborne optical depth measurements as well as those from the surface
measurements.
4. Theory of measurements: The ATSP measures direct beam solar
radiation in 6 channels in the visible and near infrared wavelengths.
The
solar radiation data are collected in the the form of voltages. The
instrument was calibrated after the experiment at the Mauna Loa
Observatory, Hawaii, using the Langley plot technique. For calibration,
data are collected at a number of solar angles from low solar elevation
(airmass=5) to high solar elevation (airmass=1.8). A regression is
developed between log voltage and airmass. This regression equation is
then extrapolated to an airmass of 0. This value is called the zero
airmass intercept voltage, and is the value that is used for calibration
of
the instrument.
The voltages measured by the instrument during Hapex-II/Sahel were
converted to total optical depth using the zero airmass intercept
voltages
calculated during the calibrations. The Rayleigh optical depth is
calculated using pressure measured on the aircraft. Finally NO2 and
ozone optical depth are subtracted from the total minus Rayleigh optical
depth. NO2 abundance was obtained from climate tables based on
Noxon, 1979 and convolved with absorption coefficients at ATSP
wavelengths. Ozone optical depth was calculated using ozone
abundances from the TOMS satellite instrument convolved with
absorption coefficients at ATSP wavelengths. The result of this
processing
is the aerosol optical depth measured in 5 channels (not including the
940
nm water vapor channel) at approximately 2 second intervals in the air
and 10 second intervals at the airport.
For atmospheric correction, the correction of remote sensing data
acquired
from satellites or aircraft for effects due to the intervening atmosphere
has proven to be a difficult problem. Not only does the atmosphere
reduce the transmission of the incoming , reflected, and emitted
radiation,
but it contributes reflected and emitted radiation of its own. Under some
conditions, atmospheric radiation comprises over 90 percent of the
satellite
observed radiance, but even much smaller effects would degrade the
quantitative use of these data unless they are taken into account.
The interaction of radiation with the atmosphere is complex and has
proved difficult to calculate without reference to measurements made at,
or close to, the time and location of interest. Effects due to Rayleigh
scattering from atmospheric gases are well understood because the major
gases (nitrogen, oxygen) which comprise 99% of the atmosphere are well
mixed and their concentrations with altitude are known. The effects due
to small particle (aerosol) scattering are quite variable due to the
wide
range of aerosol concentrations, and to the variety of aerosols found in
the atmosphere. Because aerosol concentrations cannot be known a priori,
they must be measured at the time and location of remote sensing data
acquisition.
The physical properties of aerosols such as size, shape, refractive
index,
and concentration in the atmosphere control the aerosol interaction with
light according to a set of optical properties. Three fundamental
properties are (1) the aerosol optical depth -- an indirect measure of
the
size and number of particles present in a given column of air, (2) the
single scattering albedo -- the fraction of light intercepted and
scattered
by a single particle, and (3) the phase function -- a measure of the
light
scattered by a particle as a function of angle with respect to the
direction of original propagation.
5. Equipment
5.1 The instrument consists of a solar-tracking system, detector module,
temperature control system, nitrogen-purge system, mechanical drive
chain, and data collection system.
5.1.1 The ATSP is mounted on the NASA Ames C-130 Earth Resources
Aircraft.
5.1.2 The Airborne Tracking Sunphotometer was developed for the
purpose of obtaining accurate multispectral atmospheric extinction
measurements at different altitudes. The solar-tracking system was
designed to achieve two objectives: first, to be able to acquire the sun
starting from a position several degrees away; and second, to track the
sun with an accuracy of +/-2 degrees in presence of aircraft movements.
5.1.3
The primary quantity being measured is the aerosol optical depth.
5.1.4 The sensors used are Clairex photoresitors that have been matched
to track each other over the operational range of sun intensities. The
sensing technique uses a shadow mask that bisects each detector when
the system is in balance. The design allows for very accurate tracking,
yet
at the same time provides a FOV and accurate tracking in a very compact
package. The dome rotation is referred to as azimuth motion. The central
section of the dome is free to rotate within the dome, perpendicular to
the azimuth, and is referred to as elevation motion. The control system
is designed to compensate for the flight characteristics of the aircraft.
5.1.5 The six separate detectors view the sun simultaneously at six
independent wavelengths. The FOV of each detector is set by the
entrance aperture to 4 deg, the inside surfaces of the aperture assembly
are anodized a dull black to reduce internal reflections. The 4 deg FOV
was selected to allow for +/-1 deg of tracking error without affecting
the
solar-radiation signal. The wavelengths and the full width half maximum
(FWHM) of the ATSP are shown in the following table for all channels
except for the 940 water vapor channel.
Wavelength
nm
FWHM
nm
379.8
451.3
525.7
860.5
940
1059.9
11.0
6.2
9.1
13.0
15.2
12.7
The system is designed to move in elevation or azimuth at 8 deg per
second. The acceleration that may occur during a turn is estimated to be
1.0 rad per second squared. If the instrument should lose lock, the
reacquisition occurs very rapidly as long as the sun is in the FOV of the
instrument. The tracking system responds quickly because it uses a single
rate of 8 deg per second for tracking.
5.1.6 Manufactured by NASA Ames Research Center, Moffett Field, CA
94035, Dr. Philip Russell, Principal Investigator.
5.2
Calibration
5.2.1 The detectors are temperature controlled and the amplifier gains
are
set with precision resistors. The resolution of the detector signals is
limited by the 12 bit analog-digital converter that can resolve 1 part
out
of 2048 of the 0 to +10v detector signals. The instrument is designed to
operate in clear skies and it is also assumed that over the period of a
flight profile, there are no solar fluctuations.
5.2.1.1 There is evidence in the literature that in the wavelength
region
of interest solar fluctuations would account for less than a 1% variation
of the data.
5.2.2 The instrument was calibrated at the Mauna Loa Observatory,
Hawaii from October 27 to November 5, 1992.
5.2.3 The following tables shows the calibration coefficients, corrected
for
Earth-Sun distance, calculated during the calibration at Mauna Loa
Observatory, Hawaii on October 27 to November 5, 1992 which were
used to process the ATSP data collected during Hapex-II/Sahel in August
and September, 1992.
Wavelength
nm
October 27November 5
379.8
451.3
525.7
860.5
1059.9
2.9904
4.3477
3.6111
4.0703
4.8617
6. Procedure
6.1 The data collection system was based on a Hewlett-Packard HP9816
computer with floppy disc and printer. This used data-collection,
data-processing, and printing software developed by NASA/Ames.
Besides the computer, the data collection system includes a multiplexer,
a 12-bit analog to digital converter, and electronics to process the
aircraft
inertial navigation data. The data are sampled approximately every 2 sec
and are synchronized with the aircraft data system which provides the
altitude, longitude, and latitude data. The science dataset includes the
six
detector signals, detector temperature, tracking error, sun tracker
azimuth
angle, sun tracker elevation angle, and UTC time. The computer stores
the data on 3.5 in. floppy disks. The data are also printed out for
real-time check and backup. Data were collected for the duration of the
optical flights of the C-130 aircraft, and were collected on the ground
before and/or after the flights to determine the atmospheric slab aerosol
optical depth.
6.2
Spatial Characteristics
6.2.1 The Airborne Tracking Sunphotometer has a 2 degree field of view.
The Sunphotometer measures the direct beam of solar radiation
6.2.2 See #6.2.1 above
6.3
Temporal Characteristics
6.3.1 Data were acquired during 11 days of the experiment during
optical flights of the C-130 aircraft. These days were August 29,
September 1, 2, 3, 6, 8 10, 12, 13 and 17. On September 12, ATSP data
were acquired at the airport during a SPOT overpass.
6.3.2 The data acquisition rate is 0.5 Hz. The data are sampled
approximately every 2 seconds in the air and approximately every 10
seconds on the ground.
7. Observations
8. Data Description: There are nine header records with each data file
which describe the characteristics and format of the data.
8.2 Type of data
8.2.1
8.2.2
8.2.3
8.2.4
8.2.5
Variable
Description
Range
Units
Source
Record
Number
Integer
1-5840
Integer
Data
Time
clock
Hours Minutes
0-2400
UT
C-130
Seconds
Potential
Latitude
N/A
13-14 N
Approx.
Degrees.
Decimal Minutes
C-130 Nav.
Longitude
N/A
1-4 E
Approx.
Degrees.
Decimal Minutes
C-130 Nav.
Altitude
Press./Alt.
Mean Sea
150-8,500
Meters
C-130
Level
Approx.
Solar
elevation
N/A
10-85
Approx.
Degrees
Algorithm
Airmass
N/A
1-7
Approx.
Dimensionless
Algorithm
Rayleigh
Tau 379.8nm
N/A
.150-.441
Dimensionless
Algorithm
Rayleigh
Tau 451.3nm
N/A
.074-.216
Dimensionless
Algorithm
Rayleigh
Tau 525.7nm
N/A
.039-.115
Dimensionless
Algorithm
Rayleigh
Tau 860.5nm
N/A
.005-.016
Dimensionless
Algorithm
Rayleigh
N/A
Tau 1059.9nm
.002-.007
Dimensionless
Algorithm
Aerosol
Tau 379.8nm
N/A
.046 5.096
Dimensionless
Algorithm
Aerosol
Tau 451.3nm
N/A
.092-5.552
Dimensionless
Algorithm
Aerosol
Tau 525.7nm
N/A
.064-5.39
Dimensionless
Algorithm
Aerosol
Tau 860.5nm
N/A
.065-5.566
Dimensionless
Algorithm
Aerosol
N/A
Tau 1059.9nm
.09-5.746
Dimensionless
Algorithm
Aerosol
Tau 379.8nm
uncertainty
N/A
.005-.160
Dimensionless
Algorithm
Aerosol
Tau 451.3nm
uncertainty
N/A
.003-.120
Approx.
Dimensionless
Algorithm
Aerosol
Tau 525.7nm
uncertainty
N/A
.003-.120
Approx.
Dimensionless
Algorithm
Aerosol
Tau 860.5nm
uncertainty
N/A
.003-.10
Approx.
Dimensionless
Algorithm
Aerosol
N/A
.003-.10
Dimensionless
Algorithm
Tau 1059.9nm
uncertainty
Approx.
The algorithm for processing Airborne Tracking Sunphotometer data was
written by John Livingston, SRI International and NASA Ames Research
Center and is described in Spanner et al., 1990 and Wrigley et al., 1992.
8.3 Example of data entry: The following provides an example of the data
as it has been supplied to the Hapex-II/Sahel Information System. There
is one file for each day of data collection with the Airborne Tracking
Sunphotometer. Each file has nine header records, followed by the data
collected by the Airborne Tracking Sunphotometer.
1 152804 13.552 2.857 4725 35.58 1.714 .249 .122 .065 .009 .004 0.062
0.091 0.093 0.082 0.094 .007 .006 .007 .006 .006
9. Data Manipulations
9.1
Formulas for processing data can be found in Spanner et al., 1990.
9.1.1 Description of algorithms can be found in 9.1
9.2
Data Processing Sequence
9.2.1 Briefly, the sequence begins with the voltages from each of the
six
photodetectors in the Sun photometer. It was assumed that the
attenuation of solar radiation was adequately described by the
Bouguer-Lambert-Beer extinction law:
2
V = (R'/R) Vo exp(-mt)
(1)
where V is the detector voltage, Vo is the zero air mass voltage
intercept
for the mean Earth-Sun separation R', R is the separation at the time of
observation, m is the air mass between the instrument and the Sun, and
t is the total optical depth above the sun photometer. Detector voltages
were screened to remove low values due to attenuation by clouds, loss
of Sun acquisition during steeply banked turns, or obstruction of the Sun
by the C-130 tail section. Air mass was calculated from solar ephemeris
data. Total optical spectral optical depths were calculated using
Equation
(1). The total optical depths included attenuation due to molecular
(Rayleigh) scattering, aerosol extinction, and gaseous absorption:
t
= t
+ t
+ t
+ t
+ t
(2)
t
r
a
tO3
NO2
H2O
To calculate the aerosol optical depth, the Rayleigh optical depth was
subtracted from the total optical depth. Rayleigh optical depth is
calculated from the atmospheric pressure measured on the C-130 aircraft.
Then NO2 and ozone optical depth are subtracted from the total minus
Rayleigh optical depth. The ozone abundance used to calculate ozone
optical depth was 286 Dobson units, as determined from the TOMS
satellite instrument. The following table shows the values calculated
for
NO2 and ozone optical depth which were subtracted from the Total
minus Rayleigh optical depth to derive the aerosol optical depth.
Wavelength nm
NO2 Tau
Ozone Tau
379.9
451.3
525.7
860.5
940
1059.9
0.0027
0.0023
0.0008
0.0
0.0
0.0
0.0
0.0014
0.0172
0.0008
0.0
0.0
9.2.2 The processing sequence has not changed over time.
9.2.3 No special corrections or adjustments have been made.
9.4 Graphs and Plots can be found in Spanner et al. 1990 and Wrigley
et al. 1992.
10.
Errors
10.1 The primary source of error for derivation of optical depths from
Airborne Tracking Sunphotometer data is the slowly changing set of zero
air mass voltage intercepts. The instrument is temperature stabilized
and that removes the low order voltage drifts, but the filter/detector
packages degrade slowly in time due to a variety of factors. It must be
assumed the degradation is linear in time between mountain-top
calibrations. We used a 2 percent uncertainty in the zero airmass
intercept
voltage for the 379.8 nm channel and a 1.2 percent uncertainty for the
other channels.
10.2
Quality Assessment
10.2.1 A number of side-by-side measurements of total spectral optical
depths were made with Sunphotometers from other investigators during
FIFE in order to intercompare the instruments. Bruegge et al., 1992
reported the optical depths measured in 1989 from the ATSP agreed
within +/-0.005 units of optical depth compared to 2 other field
Sunphotometers.
10.2.2
See 10.2.1 and Bruegge et al., 1992.
10.2.3
Quantitative error estimates were not performed
11.
Notes
11.1 The filter/detector combination for the 380 nm channel is known to
degrade the most quickly and might be the most susceptible to errors in
the interpolation of zero air mass intercepts with time. The aerosol
optical depth at 940 nm was not calculated because this channel primarily
measures absorption due to water vapor. We intend to calibrate the 940
channel in the future, but at this time, the channel is not calibrated
and
therefore, the water vapor overburdens were not calculated.
11.2 The values of aerosol optical depth are accurate instantaneous
values
of aerosol optical depth. These data were taken every 2 seconds. For
flightlines, if the variability is not too high, it is recommended that a
mean aerosol optical depth is calculated for the time period of the
flightline. Airborne aerosol optical depths that exceed approximately
.25
are the result of cirrus or other clouds in the solar path. Aerosol
optical
depth data obtained on the ground that exceed approximately .5 result
from cirrus or other clouds.
12. References
12.1 Matsumoto, T, P. Russell, C. Mina, W. Van Ark, and V. Banta. 1987.
Airborne Tracking Sunphotometer. J. Atmos. and Oceanic Tech. 4:336-339.
12.2 Bruegge, C.J., R. Halthore, B. Markham, M. Spanner, and R.
Wrigley. In press. Aerosol optical depth retrievals over the Konza
Prairie.
J. Geophys. Res., 97(D17):18743-18758.
King, M., D. Bryne, B. Herman, and J. Reagan. 1978. Aerosol size
distributions obtained by inversion of spectral optical depth
measurements. J. Atmos. Sci. 35:2153-2167.
Noxon, J. 1979.
84, 5067-5076.
Stratospheric NO2, 2 Global behavior, J. Geophys. Res.,
Russell, P., T. Matsumoto, V. Banta, J. Livingston, C. Mina, D. Colburn,
and R. Pueschel. 1986. Measurements with an airborne, autotracking,
external-head sunphotometer. Proc. 6th. Conf. on Atmospheric Radiation.
Williamsburg, VA, May 13-16, Am. Meteol. Soc. 4p.
Spanner, M., R. Wrigley, R. Pueschel, J. Livingston, and D. Colburn.
1990.
Determination of atmospheric optical properties for the First ISLSCP
Field
Experiment (FIFE). J. Spacecraft and Rockets, 27, 373-379.
van Heuklon, T. 1979. Estimating atmospheric ozone for solar radiation
models. Solar Energy, 22:63-68.
Wrigley, R.C., M.A. Spanner, R.E. Slye, R.F. Pueschel, and H.R. Aggarwal.
1992. Atmospheric correction of remotely sensed image data by a
simplified model, J. Geophys. Res., 97(D17):18797-18814.