New Hampshire’s Forest Stewardship Spatial Analysis Project
September 2007 (revised March 2008)
Submitted by Karen Bennett, Forest Stewardship Coordinator, UNH Cooperative Extension
Coauthored by Matt Tansey, Resource Analyst, State of NH DRED, Division of Forests & Lands
Project Background and Goals
Since 1990 the Forest Stewardship Program has been the primary mechanism to deliver landowner
assistance to forest landowners in New Hampshire. Focusing on integrated resource management and long
term planning, approximately 2300 plans have been written, representing 25% of the private lands in the
state. The New Hampshire Spatial Analysis Program (NH SAP) includes two broad areas of activity:
• Using geo-referenced information, determine the “important forest areas” in the state and rate them as
high, medium, and low.
• Digitize the property locations of forest stewardship plans and develop a geo-database that can be used
to track forest stewardship plans and accomplishments.
The project goals include:
• Spatially assess and analyze the status and distribution of existing forest stewardship plans and their
proximity to important forestlands, thus enabling forest managers to better capture and articulate the
impact of the Forest Stewardship Program.
• Provide a tool for future strategic program delivery, while demonstrating accountability for program
outcomes.
• Provide the capability to maintain Forest Stewardship Program accomplishments on an on-going basis.
The SAP analysis helps us look back on the location of past plans helping to assess program impact and to
look to the future to help direct future work of the Forest Stewardship Program.
Role of the Forest Stewardship Committee
The New Hampshire Forest Stewardship Committee advises the State Forester and UNH Cooperative
Extension as to program direction for private forest landowner assistance programs, including the Forest
Stewardship Program. The stewardship committee endorsed the Spatial Analysis Project at its annual
meeting in February 2006, provided general direction to the Spatial Analysis Project (NH SAP), and
appointed a subcommittee1 to establish and rank the attributes that should be included as part of the
important forest areas analysis. The SAP subcommittee determined which natural resource factors play the
most important role in forest stewardship in New Hampshire. A data team2 further refined what geographic
data was available and how to best use it in the analysis.
Natural Resources Included in the Important Forest Areas Analysis
The subcommittee determined the following attributes should be included in the development of the
important forest areas (listed in order of importance as determined by the committee):3
• Unfragmented Forest
• Important Forest Soils
• Significant Habitat/Natural Communities
• Proximity to Permanently Protected Land
• Aquifers
• Riparian Areas
• Proximity to Public Lands
New Hampshire’s Forest Stewardship Spatial Analysis Project
September 2007, revised March 2008
Page 1 of 17
•
•
•
•
Public Wellheads
Wetlands
Watersheds
Bedrock Geology
Initially, the subcommittee didn’t include the threats (forest health, development, wildfire risk) as required in
the national SAP protocol in the above ranking, instead choosing to focus the “important forest areas” as
“pristine” forest areas and endeavoring to use the Forest Stewardship Program as a means to maintain these
more “pristine” areas. The concern was that including threats in the analysis would cloud the identification
of the important areas.
Armed with these rankings, the data team sought geospatial data of the resources, combined some of them,
and eventually decided to include one threat- development pressure- in the analysis. Notably aquifers and
public wellheads were combined into “drinking water protection areas” and bedrock geology was reflected
in important forest soils. The forest health risk assessment was changed into a “healthy forest ratings”.
Though fire risk was also called for in the protocol, spatial data exists, and according to other Forest Service
analyses, New Hampshire has the highest percentage of homes in the wildland urban interface; fire wasn’t
included because the subcommittee felt it didn’t act as an appropriate filter for this analysis.
Unfragmented Forests (Forest Block Size): Unfragmented forest was
selected as the single most important resource characteristic by the NH
SAP committee. New Hampshire Fish and Game has developed a data
layer that quantifies blocks of forest with a series of metrics. It is based on
the 2001 Land Cover Assessment provided by GRANIT4. Blocks were
given a rank based on size. 50 to 100 acre blocks received 1 point, 100 to
500 acre blocks received 2 points, 500 to 1000 acre blocks received 3
points and blocks over 1000 acre received 4 points.
Important Forest Soils Groups: Highly productive forest soils were the
second most important resource factor according to the committee. All
spatial soils data was produced by the Natural Resources Conservation
Service (NRCS) and soils types were joined to their respective Important
Forest Soils Group (IA, IB, IC, IIA, IIB)5. The Important Forest Soils
Group classification takes into account soil productivity for forestry but
also slope and operability. IA soils represent the best hardwood sites.1B
represents slightly less fertile hardwood sites and group. IC represents
sandy outwash sites that favor softwood growth. After a first draft IB soils
were excluded because they fell too broadly over the landscape. Group IA
and IC were given 4 points in the final analysis to reflect their importance.
New Hampshire’s Forest Stewardship Spatial Analysis Project
September 2007, revised March 2008
Page 2 of 17
Significant Habitat (Wildlife Habitat Data): The NH SAP committee
ranked significant habitat which includes important wildlife habitat and
natural communities almost as high as important forest soil groups. NH
Fish and Game created a GIS analysis to rank wildlife habitat within the
state for their Wildlife Action Plan.6 This analysis referenced the natural
communities as designated by the NH Natural Heritage Bureau.
Numerous models and techniques were used to create several 30 meter
resolution grids. The wildlife habitat tier data set was selected for use in
the NH SAP. Tier 1 represents the highest quality habitat in the state, tier
2 represents the highest quality habitat in the biological region (defined
by TNC ecoregional subsection for terrestrial habitats) and tier 3
represents the supporting landscape. These tiers were assigned 3, 2, 1
points respectively. For more information on the New Hampshire Fish
and Game Departments Wildlife Action Plan visit:
http://www.wildlife.state.nh.us/Wildlife/wildlife_plan.html
NH Fish and Game created a GIS analysis to rank wildlife habitat within
the state for the NH Wildlife Action Plan (WAP).7 A statewide wildlife habitat landcover of 16 habitat types
was created for the WAP. The habitats were then analyzed for condition using biological, landscape and
human influence factors that inform habitat condition. Thresholds were set to determine the highest ranking
condition for each habitat type statewide, or Tier 1 (the top 10-15% by area of each). Since the state is
climatically and geologically diverse, the habitats were then ranked within each biological region,
(ecoregional subsection or watershed grouping) to create Tier 2. A third tier, Supporting Landscapes,
consists of large forest blocks and the upland area of high ranking watersheds, plus known occurrences of
some rare wildlife and natural communities. The Wildlife Habitat Condition Rankings were selected for use
in the NH SAP.
Proximity to Permanently Protected Lands (Private Conservation
Lands): The state conservation land data layer maintained by GRANIT
tracks both public and private conservation land. It was used to identify
private conservation land. This data set is the most comprehensive
collection of undeveloped parcels over two acres that are protected from
future development. The data is complete for the state and is updated
regularly. Parcels over 50 acres were buffered by 1000 feet and the
buffer received 1 point. The actual parcel received 3 points. Other
conservation parcels under 50 acres received 2 points but were not
buffered.
New Hampshire’s Forest Stewardship Spatial Analysis Project
September 2007, revised March 2008
Page 3 of 17
Riparian Areas: This data layer supplied by NH Fish and Game includes
all perennial streams, lakes and ponds and hydrologically connected
wetlands. All elements are buffered 300 feet and get one point.
Proximity to Public Lands (Public Conservation Land): The state
conservation land data layer maintained by GRANIT was also used to
delineate all fee owned public land. This includes all land held by federal,
state, county and towns. Several other large parcels, such as Jericho
Mountain State Park, that were not in the original data layer were added
provided there was spatial data available. Parcels over 50 acres were
buffered by a 1000 feet and received one point.
Aquifers and Public Wellheads (Drinking Water Protection Areas):
Two data layers were used to create a composite of areas that are
important to drinking water protection. The well head protection areas
layer, supplied by NH Department of Environmental Services, locates
drinking water supply areas and puts a protective buffer around the point
the size of which depends on the numbers of people served. This was
combined with a high transmissivity aquifers data layer which was
produced by USGS. These two layers were combined into one and all
areas of interest were given one point.
New Hampshire’s Forest Stewardship Spatial Analysis Project
September 2007, revised March 2008
Page 4 of 17
Wetland Areas: The National Wetlands Inventory contains the location
and classification of wetlands as defined by the U.S. Fish & Wildlife
Service. This data layer was queried to select only palustrine wetlands. No
buffers were used and all areas with wetlands were given one point.
Watershed Quality: The SPARROW8 (Spatially Referenced Regression of
Contaminant Transport on Watershed attributes) data layer produced by
USGS was used to assess watershed quality. This research indexes the
quality of all stream catchments in the state based on regression models of
total nitrogen and total phosphorus transport. The data is based on stream
reaches which has a much finer resolution than the HUC 12 watersheds. The
top 25% of stream catchments, with an index value of 1 – 1.015584, were
chosen to represent the best quality watersheds across the state and assigned
one point.
Healthy Forests: Annually, the Forest Health Section of the NH Division of
Forests and Lands flies the state in fixed wing aircraft and creates a data
layer of forest damage in accordance with USDA Forest Service standards.
This data has been used to produce an analysis of forest health from 1990 to
2006 for all of New Hampshire excluding the White Mountain National
Forest. This particular analysis is a 30 meter grid layer. The original protocol
for the SAP called for the identification of areas that have sustained forest
damage. The NH SAP committee felt it was more appropriate to select areas
that were less damaged and more “healthy”. To this end areas of forest that
were damaged 3 times or less over since 1990 were given a value of one
point in the analysis.
New Hampshire’s Forest Stewardship Spatial Analysis Project
September 2007, revised March 2008
Page 5 of 17
Change in Housing Density: The greatest threat to forest land in New
Hampshire is development. Data from David Theobald’s research on
housing density9 was used to determine the areas in the state that are under
threat from development. This is a 30 meter raster based on 2000 US
Census Bureau block (SF1) datasets. A growth model was used to predict
the growth in housing units on the same census block while subtracting land
that was non developable (water, transportation infrastructure or protected
land). The NH SAP committee felt that land under the highest threat to
development was beyond the reach of the Forest Stewardship Program and
land under the least threat not as needy of attention so census blocks that
showed an intermediate rate of growth were selected for areas in this
analysis and given one point.
Masked Areas: The Forest Stewardship SAP excludes several areas that don’t qualify for inclusion in the
forest stewardship program. All publicly fee owned lands (federal, state, town, county) are excluded. This
diversity of properties was queried from the conservation lands data layer in GRANIT. The 2001 Land
Cover Assessment10 was used to isolate areas of residential, commercial, or industrial development,
transportation, and open water, which were also used in the mask.
The protocol called for an additional mask to be created for areas that were in the analysis area but were non
forest and non-developed. The 2001 Land Cover Assessment was used to isolate areas of row crops, hay or
pasture, open wetlands, tidal wetlands, gravel pits or quarries, bedrock, sand dunes and tundra. Tundra was
included in the mask but will make no difference in the analysis because all tundra land cover is on state or
federal land and already masked out. Areas of land cover that were described as orchard, forested wetland
and other cleared, which is land recently harvested, were associated with forestland.
Weighting Rationale and Final Outcome of the Important Forest Areas Analysis
After the data layers for this project were selected and discussed, the subcommittee then voted on the
perceived importance of each layer (see footnote 1). The votes were tallied and could have been used to
determine a weighting scheme between the layers however; it was decided to give weighting within specific
data layers.11 As an example, forest block size was considered the most important factor and that as the size
of the forest block increases it becomes more important and should receive a higher score. So a score of one
through four was applied to each forest block. Similarly, wildlife habitat was given a score up to three to
reflect the importance of different habitat quality. Other data layers like the palustrine wetlands were
uniformly given a value of one point regardless of size, composition or other factors. Taking a real example
from the data; if a pixel fell on the largest forest block (4 points), had the highest habitat rating (3 points),
was permanently protected (3 points), and had important forest soils (4 points), it would have a score of 14
points and rise toward the top of the ranking. Conversely, if a pixel fell on a riparian buffer (1 point), in
healthy forest (1 point), and occurred in a priority watershed (1 point) it would only have a score of 3 points
and be near the bottom of the overall classification.
New Hampshire’s Forest Stewardship Spatial Analysis Project
September 2007, revised March 2008
Page 6 of 17
After the processing was completed and the data layers were rasterized the raster calculator was used to
create a final overlay. The standard for this project was to divide the final composite into values of high,
medium and low using natural breaks in the data (Jenks method)12. Of the area considered in the analysis 1.7
million acres (39%) were considered to have high stewardship program potential, 1.8 million acres (41%)
were considered medium, and 841 thousand acres (19%) were considered to have low program potential.
Table 1. Amount of Land with High, Medium, and Low Stewardship Potential
Developing the Forest Stewardship Plan Data Layer
Forest stewardship plans written since the start of the program through fiscal year 2005 were digitized and a
data layer of plans was created. Each landowner was assigned a unique identifying number and this data
layer was referenced to an access database containing contact information and acres covered by the plans. A
total of 591,111 acres of plans (14% of the NIPF acres in the state) were digitized, just over half of the plans
written since the start of the program. Some of the plans not digitized couldn’t be found, were out of date,
were written after the project started, or had other reasons for not being included in this data layer. This data
layer and database will be maintained and improved. Future work will include recovering and digitizing
relevant older plans, digitizing and adding plans written after FY 2005, and developing a protocol for
digitizing new plans and capturing landowner information to a centralized geo-referenced database.
The forest stewardship plan data layer was combined with the important forest areas analysis to create a set
of maps that follow. These maps are:
Map 1: Potential for Forest Stewardship Benefits
Map 2: Potential for Forest Stewardship Benefits and Existing Stewardship Plans
Map 3: Forest Stewardship Potential on Private Lands and Existing Stewardship Plans
Map 4: Resource Richness
Map 5: Resource Threats
Map 6: Forest Stewardship Potential on Non-Forested – Non-Developed Lands & Existing Stewardship Plans
Map 7: Regional Map – Sullivan County
Table 2 shows that, 32% of the lands with high stewardship potential are covered by plans. These maps will
be used to direct future programming, for example highlighting lands with high potential for the stewardship
program, enabling us to do education and outreach to landowners in these areas.
Table 2. Plans According to Stewardship Potential
New Hampshire’s Forest Stewardship Spatial Analysis Project
September 2007, revised March 2008
Page 7 of 17
New Hampshire’s Forest Stewardship Spatial Analysis Project
September 2007, revised March 2008
Page 8 of 17
New Hampshire’s Forest Stewardship Spatial Analysis Project
September 2007, revised March 2008
Page 9 of 17
New Hampshire’s Forest Stewardship Spatial Analysis Project
September 2007, revised March 2008
Page 10 of 17
New Hampshire’s Forest Stewardship Spatial Analysis Project
September 2007, revised March 2008
Page 11 of 17
New Hampshire’s Forest Stewardship Spatial Analysis Project
September 2007, revised March 2008
Page 12 of 17
New Hampshire’s Forest Stewardship Spatial Analysis Project
September 2007, revised March 2008
Page 13 of 17
New Hampshire’s Forest Stewardship Spatial Analysis Project
September 2007, revised March 2008
Page 14 of 17
For More Information:
Karen P. Bennett, Extension Professor & Specialist, Forest Resources
UNH Cooperative Extension, 212 Nesmith Hall, 131 Main St., Durham, NH 03824
(603)862-4861, karen.bennett@unh.edu
Matthew B. Tansey, Forest Resource Analyst
State of NH DRED, Division of Forests & Lands, PO Box 1865, Concord, NH 03302
(603)271-2214, mtansey@state.nh.us
Acknowledgements:
Chuck Hersey, Sullivan County Extension Forester, and Sam Stoddard, Coos County Extension Forester
provided leadership and direction for the forest stewardship plan digitizing aspect of this project. Other
foresters contributing significantly to the success of this project include: Tim Fleury, Jon Nute, Northam
Parr, Steve Roberge, and Wendy Scribner. Deb Stevens, Administrative Assistant, Rockingham County
Extension also contributed.
1
SAP subcommittee included: Karen Bennett, Ann Davis (landowner), Chuck Hersey (UNH Cooperative Extension), Roger
Monthey (Forest Service), Wendy Scribner (UNH Cooperative Extension), Jim Spielman (NRCS), Matt Tansey (NH Division of
Forests and Lands)
2
Data layer team: Karen Bennett, Chuck Hersey, and Sam Stoddard, (UNH Cooperative Extension), Tom Luther and Roger
Monthey (Forest Service), Matt Tansey (NH Division of Forests and Lands)
3
Resource factors chosen by the subcommittee to be included in the important forest areas analysis as ranked by importance. The
lower the number, the higher the ranking
Unfragmented Forest- 11
Important Forest Soils- 25
Significant Habitat/Natural Communities- 27
Proximity to Permanently Protected - 33
Aquifers- 37
Riparian- 42
Proximity to Public Lands- 46
Public Wellheads- 52
Wetlands- 56
Watersheds- 61
Bedrock Geology - 73
4
The New Hampshire Geographically Referenced Analysis and Information Transfer System (NH GRANIT)
http://www.granit.sr.unh.edu/ is a cooperative project to create, maintain, and make available a statewide geographic data base
serving the information needs of state, regional, and local decision-makers. A collaborative effort between the University of New
Hampshire and the NH Office of Energy and Planning, the core GRANIT System is housed at the UNH Institute for the Study of
Earth, Oceans, and Space in Durham. It includes a geographic database, hardware and software to build, manage, and access the
database, and a staff of experts knowledgeable in geographic information systems, image processing, and computer analysis. In
addition to database development and maintenance, the GRANIT staff offers a range of application development, training, and
related technical services to GIS users in the state and the region.
The GRANIT approach to a statewide GIS depends upon the cooperative efforts of a host of agencies, collaborating on various
elements of database design and construction as well as application development. The collaboration occurs formally through the
NH GIS Advisory Committee, and informally through daily interactions between the growing body of GIS users in the state and
the region.
New Hampshire’s Forest Stewardship Spatial Analysis Project
September 2007, revised March 2008
Page 15 of 17
5
Important Forest Soils
Important Forest Soil Group IA
Soils belonging to this group consist of the deeper, loamy textured, moderately well, and well-drained soils. Generally, these soils
are more fertile and have the most favorable moisture relationships. Forest successional trends on these soils are toward stands of
shade tolerant hardwoods, usually beech and sugar maple. Hardwood competition is severe on these soils. Softwood regeneration
is usually dependent upon persistent hardwood control efforts. On this soil type sugar maple is favored by selection cutting
methods; white ash and yellow birch are favored by group and strip cutting; white ash is favored by shelterwood cutting; and
white birch is favored by clearcutting.
Important Forest Soil Group IB
Soils assigned to Group IB are generally sandy or loamy over sandy textures and slightly less fertile than soils in Group IA. Soil
moisture is adequate for good tree growth, but may not be quite as abundant as in Group IA soils. Forest successional trends on
these soils are toward shade tolerant hardwoods, predominantly beech. Hardwood competition is moderate to severe on these soils.
Successful softwood regeneration is dependent upon hardwood control. On Group IB soils white birch is favored by clearcutting,
yellow birch is favored by group and strip cutting; hemlock and red spruce are favored by selection cutting; and white pine is
favored by shelterwood cutting.
Important Forest Soil Group IC
The soils of this group are outwash sands and gravels. Soil drainage is excessively drained to moderately well-drained. Soil
moisture is adequate for good softwood growth, but is limited for hardwoods. Forest successional trends on these coarse-textured,
somewhat droughty and less fertile soils are toward stands of shade tolerant softwoods, especially red spruce and balsam fir.
Balsam fir is a persistent component of stands on this soil type, but is shorter lived than red spruce. Hardwood competition is
moderate to slight on these soils. Due to less hardwood competition, these soils are ideally suited for softwood production. With
modest levels of management, white pine can be maintained and reproduced on these soils. Because these soils are highly
responsive to softwood production, they are ideally suited for forest management. On these soils white pine is favored by group
and strip cutting, or shelterwood cutting; red spruce and balsam fir are favored by selection cutting or shelterwood cutting; and
hemlock is favored by selection cutting.
6
Information about habitat condition was analyzed to develop a statewide and regional ranking and identify the highest quality
habitat relative to all polygons of a given habitat type in the state. Tiers were determined by assigning threshold values to each
habitat type. The Tier 1 rating was given to areas that contain the highest quality in the state. Tier 2 areas contain the highest
quality in the biological region (defined by the TNC Ecoregional Subsection for terrestrial habitats or the Watershed Group for
wetland and aquatic habitats), Tier 3 includes other significant factors such as the entire watershed of high quality stream and
lakes, large forest blocks of statewide significance, or specific animal, plant and natural community occurrences identified by
NHFG as critically imperiled or by NHNHB as highest importance. Each habitat polygon has attributes showing the condition
scores so that local communities can set their own condition thresholds for habitats of local importance. Results will be re
evaluated to monitor the effectiveness of conservation actions and respond appropriately to new information or changing
conditions.
7
For more information on the New Hampshire Wildlife Action Plan: http://www.wildlife.state.nh.us/Wildlife/wildlife_plan.htm
Emily Brunkhurst, Emily.P.Brunkhurst@wildlife.nh.gov contributed to this description.
8
For more information on the SPARROW model for New England see Moore et al. Estimation of Total Nitrogen and Phosphorus
in New England Streams Using Spatially Referenced Regression Models
SPARROW Surface Water-Quality Modeling Nutrients in Watersheds of the Conterminous United States Methods and Selected
Results: Smith, R.A., G.E. Schwarz, and R.B. Alexander, 1997, Regional interpretation of water-quality monitoring data, Water
Resources Research, v. 33, no. 12, pp. 2781-2798 http://water.usgs.gov/nawqa/sparrow/wrr97/results.html
9
Theobald, D.M. 2005. Landscape patterns of exurban growth in the USA from 1980 to 2020. Ecology and Society 10(1): 32.
New Hampshire’s Forest Stewardship Spatial Analysis Project
September 2007, revised March 2008
Page 16 of 17
10
Numerical designation of land types in the 2001 Land Cover Assessment in GRANIT:
residential, commercial, or industrial development (110),
transportation (140)
and open water (500)
row crops (211)
hay or pasture (212)
open wetlands (620)
tidal wetlands (630)
gravel pits or quarries (710)
bedrock (720)
sand dunes (730)
tundra (800)
orchard (221)
forested wetland (610)
other cleared (790)
11
Points assigned to natural resources
Forest block size
50- 100 acres
100- 500 acres
500- 1000 acres
over 1000 acres
Important forest soils
Significant Wildlife Habitat
Tier 1
Tier 2
Tier 3
Proximity to private conservation lands
Over 50 acres plus buffer
Under 50 acre
no buffer
Protected land
Riparian areas
Proximity to public lands
Drinking water protection areas
Wetlands
Watershed quality
1
2
3
4
4
3
2
1
1
2
3
1
1
1
1
1
12
Classes are based on natural groupings inherent in the data. ArcMap identifies break points by picking the class breaks that best
group similar values and maximize the differences between classes. The features are divided into classes whose boundaries are set
where there are relatively big jumps in the data values.
New Hampshire’s Forest Stewardship Spatial Analysis Project
September 2007, revised March 2008
Page 17 of 17