2. STUDY AREAS
2.1.
LARGE-REGIONAL STUDY AREA
2.1.1.
Land Cover and Hydrography
2.1.2.
Forest Cover
2.1.3.
Soils
2.1.4.
Precipitation and Climate
2.1.5.
Large-Regional Data Collection
2.2.
SMALL-REGIONAL STUDY AREA
2.3.
MESO-CELL STUDY AREAS
2.3.1.
North Park MSA
2.3.2.
Rabbit Ears MSA
2.3.3.
Fraser MSA
2.3.4.
MSA Data Collection
2.4.
INTENSIVE STUDY AREAS
2.4.1.
Description
2.4.2.
ISA Data Collection
2.5.
LOCAL-SCALE OBSERVATION SITE
2.6.
OTHER RESEARCH SITES/FACILITIES
2.6.1.
Niwot Ridge Biosphere Reserve and LTER
2.6.2.
Loch Vale Watershed
2.6.3. Storm Peak Laboratory
2. STUDY AREAS
The experimental design is based on a 5-level set of nested study areas ( Figure 1 ) that range from a small, 1-ha study plot to a large 3.5°x 4.5°(approximately 400-km x 400-km) region. Intermediate nested areas include nine 1-km2 sites, three 25-km x 25-km sites, and a small 2.5Ex1.5E (approximately 215-km x 170-km) region. Each nested level serves a specific purpose in meeting the overall objectives of the experiment. Activities at the smallest site will be focused on ground-based microwave remote sensing and intensive monitoring of the snow and soil conditions viewed by the remote sensing instruments. The 1-km2 sites each represent a relatively homogeneous, major physiographic characteristic. Activities at these sites will be focused on intensive spatial sampling of snow and soil properties coincident with aircraft observations. Aircraft observations will be conducted to provide full microwave remote sensing data coverage over each of the three 25-km x 25-km sites, which correspond in scale to the resolution of current spaceborne passive microwave sensors and are commensurate with typical resolutions of current meso-scale atmospheric models. The 2.5°x1.5° area is suitable for high-resolution (e.g. # 1-km) regional modeling exercises. The largest area is suitable for coarser-resolution regional and meso-scale modeling exercises. Each intermediate level is intended to provide a "scale bridge" between the levels above and below, with the explicit purpose of helping to extend understanding of processes, measurements, and models at one scale to other scales.
Two coordinate systems will
be used for the experiment. For convenience and portability of data sets
for larger regions, an un-projected latitude/longitude coordinate system
based on the WGS84 datum will be used. For smaller areas (1-km2
and 1-ha sites) where ground data are to be collected, a Universal Tranverse
Mercator (UTM) projection (Zone 13 North) based on the WGS84 datum will
be used (
Table 1
).
2.1. LARGE-REGIONAL STUDY AREA
The largest study area for the experiment ( Figure 3 ) is located in northern Colorado and southern Wyoming, U.S.A. (104°-108.5° W, 38.5°-42° N). This 4.5°x 3.5° region was selected for the experiment because:
2.1.1. Land Cover and Hydrography
The large-regional study
area consists of three major land cover regimes distributed east-west across
the area: 1) prairie grasslands and croplands in the eastern fourth of
the area, 2) desert and high-altitude plateaus in the western fourth, and
3) a mixture of forested mountains and valleys, and alpine areas in the
middle two-fourths of the area. Two large, open parklands lie within the
mountainous part of the study area: North Park, surrounding Walden, CO,
and South Park, southeast of Leadville, CO. Both of these are broad, open,
rolling grasslands at high elevation (North Park lies at ~2400-m, South
Park at ~2700-m). The large-regional study area contains the headwaters
for seven major river systems: 1) North Platte River, 2) South Platte River,
3) Arkansas River, 4) Yampa River, 5) White River, 6) Gunnison River, and
7) Colorado River. The total downstream water yield from these basins is
directly proportional to the mountain snow pack.
Forest cover types within
the large-regional study area are predominantly Spruce-Fir, Lodgepole Pine,
Ponderosa Pine, Pinyon-Juniper, and Chaparral. About one-third is unforested.
Forest cover density generally increases with elevation to the local tree-line.
Surface soils (upper 5 cm)
in the study area are predominantly sandy loam or loam. Clay soils are
prevalent in the Colorado Piedmont area (i.e. the Denver/Boulder area)
and along parts of the Colorado River. Sandy soils are found in the northeast
(e.g. South Platte valley) and northwest parts of the study area (e.g.
western Yampa valley).
2.1.4. Precipitation and Climate
Most of this region's annual
precipitation occurs as winter snowfall. Pacific frontal systems bring
most of the winter moisture to this region. These storm systems can arrive
into the region from either the west, northwest, or southwest, and this
influences the distribution of precipitation. Westerly tracks are orographically
uplifted to some extent by the Wasatch Plateau east of the study area in
Utah, and are lifted further by the ranges along the Continental Divide
in the central part of the study area. This results in the heaviest precipitation
west of the Continental Divide. Northwesterly tracks are lifted by the
Wasatch Range and the Uinta Mountains in Utah and by ranges along the divide
in the north central part of the study area, resulting in heavier precipitation
at these locations. Storm tracks arriving from the southwest don't encounter
major orographic effects until they reach the San Juan Mountains in southwestern
Colorado, just south of the large-regional study area. Heavy winter precipitation
occurs in this part of the region from these storm tracks. In general,
precipitation declines markedly throughout areas east of the Continental
Divide. However, low pressure systems east of the Divide can bring significant
moisture in from the Gulf of Mexico during Spring, resulting in sometimes
heavy snow fall in the foothills at lower elevations on the eastern side
of the Divide ("upslope" conditions). Lower elevation areas of the Central
Rockies receive considerably less precipitation; most of the region's snow
pack storage is concentrated in the higher mountains. The mean date of
snow cover formation ranges from October 15 near the Continental Divide
to November 15 for most of the rest of the study area. The mean date of
snow cover disappearance ranges from early March in the western and eastern
parts of the study area, to May 1 in all but the highest elevations near
the Continental Divide.
2.1.5. Large-Regional Data Collection
The three major data sets
to be collected for the large regional area are 1) spaceborne optical and
microwave remote sensing measurements, 2) meso-scale and regional atmosphere
and surface analyses, and 3) geographic information system (GIS) data sets
describing a variety of physiographic and logistically relevant information.
These data sets are described in detail in subsequent sections. All of
the remaining study areas for the experiment are nested within this region.
2.2. SMALL-REGIONAL STUDY AREA
The small-regional study
area (
Figure 3,
Figure 4
) is located in north-central
Colorado (105°-107.5°
W, 39.5E-41E
N), and is approximately 215-km x 170-km. This area is large enough to
a) contain examples of most of the physiographic characteristics of the
large-regional study area, and b) contain 20-80 mesoscale model grid cells,
but it is also small enough to permit efficient, high-resolution modeling
exercises. Data collection for this area will be essentially the same as
for the large-regional area. However, some high-resolution satellite remote
sensing data sets with small image areas may be collected only for this
area and not for the large-regional area.
Nested within the small-regional
study area are three 25-km x 25-km study areas that will be the focus of
airborne data collection (
Table 2
). The size
of these areas is typical of the resolution of current spaceborne passive-microwave
sensors and of meso-scale atmospheric-model grids, hence they are called
"Meso-cell Study Areas" (MSA). The primary objective in selecting these
areas was that each individual area should represent a distinct cold-region
physiographic regime (from a meso-scale perspective), while together the
three areas should represent as broad a range of regimes as possible. Topography,
forest cover, and snow characteristics were the three major criteria for
selecting these areas (
Table 3
). These three
MSAs represent four of the major global snow cover classes of Sturm
et al., [1995], which together comprise 88% of the seasonally snow
covered areas of the Earth. The remaining 12% is the maritime snow class,
which is not typically found in the experiment study region.
North Park is a broad, high-elevation
parkland approximately 40 km in diameter. The MSA is centrally located
in North Park (
Figure 5
). It has a mean elevation
of 2499-m (
Table 4
). Most of the MSA has very
low relief. It has a total elevation range of 312-m, due largely to the
presence of low foothills in the southeastern part if the MSA. North Park
is an inter-mountain glacial basin that opens north into Wyoming, and is
surrounded by the Park Range on the west, the Medicine Bow Range on the
east and northeast, the Rabbit Ears Mountains on the south, and the Never
Summer Range on the southeast. These surrounding high mountain areas develop
deep snow packs in winter as a result of significant orographic precipitation
effects, but relatively little snow accumulates in North Park itself due
to the precipitation "shadow" caused by the surrounding mountains (
Table 6
). This area forms the headwaters of the North Platte River. Several
small meandering rivers drain the surrounding mountains and flow through
the flat topography of North Park, where they join the North Platte River
just north of Walden. Consequently, parts of the MSA are relatively wet.
The MSA includes the 20,000 acre Arapaho National Wildlife Refuge with
significant riparian and wetland areas. The MSA has very little forest
cover (
Table 5
). Most of the vegetation is
sage-grassland, with willow along riparian areas. Snow packs in this area
tend to be shallow and windblown, and are typical of prairie and arctic-
and alpine- tundra snow covers (53% of the global seasonal snow cover [Sturm
et al., 1995],
Table 3
). The MSA is crossed
by two primary highways and several secondary roads that remain open during
winter, providing easy access throughout the area. The town of Walden,
located near the center of the MSA, is the principal source for services
in the area.
The Rabbit Ears MSA straddles
Rabbit Ears and Buffalo Passes over the Continental Divide, between the
Gore Range to the south and the Park Range to the north (
Figure 6
). The mean elevation of the MSA is 2725-m (
Table 4
). The topography consists primarily of low to moderately rolling
hills extending north-south throughout the middle of the MSA. These are
flanked by much flatter valleys at lower elevations along the eastern and
western edges of the MSA. The landcover of this MSA is the most diverse
of the three (
Table 5
). Forest cover within
the MSA consists of mixed stands of spruce-fir, lodgepole pine, aspen,
and scrub oak forest interspersed with broad glades and meadows. The area
receives heavy snowfall during the winter and spring (
Table 7
), and often develops the deepest snow packs in Colorado at the
higher elevations near Buffalo Pass. The moderate topography and forest
cover, along with large snow accumulation, make this a popular winter recreational
area. The MSA contains four National Resource Conservation Service (NRCS)
SNOTEL sites and five NRCS snow course sites, providing a long historical
record of snow conditions in the area. Snow packs in this area have depths
and other characteristics found in regions where orographically induced
precipitation plays a dominant role (10% of the global seasonal snow cover
[Sturm et al., 1995],
Table 3
). The
MSA is crossed east-west by Highway 40 over Rabbit Ears Pass. Other primary
and secondary roadways provide easy access throughout the southern half
of the MSA. The northern half of the MSA is less accessible, and will require
snow machines for efficient travel. The town of Steamboat Springs, along
the western edge of the MSA, is the principal source for services in the
area.
The Fraser MSA is a topographically
complex area, with a large north-south topographic gradient (
Figure 7
). Its mean elevation is 3066-m, with a range of nearly 1400-m
and a maximum elevation of 3962-m (
Table 4
).
The Continental Divide winds through high-elevation alpine areas of the
Front Range, along the southern margin of the MSA. The Fraser River and
its tributaries (Vasquez Ck., St. Louis Ck., and Crooked Ck.) flow northward
from the Divide through a finely dissected series of ridges and glacial
valleys throughout the southern half of the MSA. The northern half of the
MSA has much lower relief. Here, the Fraser River valley broadens into
a wide glacial outwash plain. Vegetation at the low elevations to the north
consists of irrigated grasslands and sparse, mixed forests (
Table 5
). The highest-elevation areas in the south are predominantly alpine
tundra or bare rock. The montane areas in between are dominated by dense
coniferous forest, generally with spruce-fir forests on the wetter north-facing
slopes and lodgepole pine on the drier south-facing slopes. The MSA contains
the Fraser Experimental Forest (FEF), a USDA Forest Service research facility
that includes the well-studied St. Louis Creek basin above Fraser, CO.
The 93-km2 St. Louis Creek basin is the largest gaged basin
fully contained within an MSA. Historical daily streamflow records are
available from seven gages within the basin. Forests within this facility
have been intensively mapped, with individual stems mapped throughout some
areas. The FEF also offers several facilities useful to the experiment,
including on-site housing, kitchen facilities, a wet-lab, office, computer
space, workshop, and over-snow vehicles. The Fraser MSA is typically cooler
than either the Rabbit Ears or North Park MSAs, due mainly to its higher
elevation (
Table 8
). The area contains two
NRCS SNOTEL sites and three snow courses, providing a long historical record
of snow conditions in the area. Snow packs in this area typically display
only minimal modification by wind and considerable depth hoar development
(25% of the global seasonal snow cover [Sturm et al., 1995],
Table 3
). The eastern two-thirds of the MSA is easily accessible
via Highway 40, which crosses Berthoud Pass over the Continental Divide
in the southeast corner of the MSA, and transects the MSA northward, following
the Fraser River. Several secondary roads in this part of the MSA also
remain open during winter, including the access road to FEF. The western
third of the MSA is less accessible. The towns of Fraser and Winter Park
are the principal sources for services in the area.
The MSAs will be the focus
of airborne data collection in the experiment. Airborne active (AIRSAR
and POLSCAT) and passive (PSR and AESMIR) microwave data will be collected,
with complete spatial coverage of each MSA (except for POLSCAT, which will
have 6 1-km wide tracks measured across the full width of each MSA). Airborne
measurements of terrestrial and atmospheric gamma radiation data will also
be collected over each MSA to measure snow water equivalent. Complete spatial
coverage of the gamma measurements is not possible, however intensively
spaced flight lines throughout each MSA will enable the best possible estimate
of the mean snow water equivalent for the MSA. This in turn will be used
for evaluation of remote sensing algorithms and land surface models. Details
of the airborne data collection are given in later sections.
Nested within each of the three MSA are three 1-km2 Intensive Study Areas (ISA), for a total of nine ( Table 9 ), (Figures 8-36). The primary objective in selection of these areas was to represent a broad range of major physiographic characteristics with a limited set of homogeneous study sites. In this sense, "homogeneous" should not be confused with the concept of "pure pixels". Each site contains a mixture of different landscape elements, but these elements are relatively consistent throughout the site and the site can be characterized by this homogeneous mixture of landscape and topographic characteristics ( Table 10, Table 11 ). Factors potentially affecting microwave response were also considered, such as the relative wetness or dryness of the site, and the type, density and orientation of forest cover. Consequently, the four sites contained within an MSA represent four major landscape (i.e. topography, vegetation, and snow cover) categories within the MSA, and also represent different anticipated "microwave response units". Site accessibility (distance to roads, land ownership, etc.) was also an important consideration in selecting the specific location of the ISA. Together, the nine sites provide a cross-section of different physiographic characteristics typically found in cold regions. This is important for addressing several objectives of the experiment, such as developing global maps of confidence and performance limits for retrieval algorithms and land surface models.
Each ISA is 1000-m x 1000-m
in size. These dimensions were chosen for several reasons. First, in ISAs
with more complex terrain, 1-km2 tended to be the largest area
for which contiguous, homogeneous characteristics could be found. Second,
this size is an important "data resolution threshold" for many land surface
modeling efforts for large regions. Several spatial geophysical data sets
that are currently available for large regions (e.g. continental U.S.),
including digital elevation, land-use/land-cover, vegetation, soils, and
moderate-resolution optical remote sensing data (e.g. AVHRR, GOES, MODIS),
share a common nominal resolution of 1-km2. For some geophysical
variables this is also the highest available resolution. Third, considering
the importance of selecting as many sites as possible for representing
a broad range of physiographic categories, this area was considered to
be the largest that could be sampled intensively based on expected resources
and logistical considerations.
At these nine sites, snow
and soil characteristics will be sampled intensively using a geostatistical
framework, whereby 1) the mean and variance of the 1-km2 site
can be accurately determined, 2) structural analysis can be performed to
evaluate and model spatial variability, and 3) the spatial distribution
of snow and soil characteristics can be estimated. A stratified-random
sampling frame with 100-m grid spacing will be used, with sample locations
located randomly within each grid cell. Each ISA will contain a meteorological
station (in Year 2 of te experiment), located near the center of the site,
to provide observations to force a variety of spatially distributed, high-resolution,
snow/soil energy- and mass-balance land-surface models, which can in turn
be evaluated with the ground observations and geostatistical analyses.
Details of the ground data collection for the ISA are given in a later
section.
2.5. LOCAL-SCALE OBSERVATION SITE
The final and smallest study
site is a single, 1-ha study site located within the Fraser ISA, near the
Fraser Experimental Forest Headquarters Facility. At this 100-m x 100-m
site, intensive ground observations of snow, soil, and vegetation will
be made in conjunction with stationary, ground-based microwave remote sensing
(active and passive) and micrometeorological observations. The site consists
of two open meadows separated by a stand of short trees. Ground-based remote
sensors will have an opportunity to view both a homogeneous area of open
snow cover, and a homogenous stand of forest cover. At this scale, where
spatial variability is relatively low, microwave remote sensing data, radiative-transfer
models, detailed physical models of energy and mass balance of snow and
the underlying soil, and ground observations can be most easily related
to each other. Here, physical models and remote sensing retrieval algorithms
can be evaluated with little ambiguity, and confidence levels can be established
at the scale of our current understanding of important physical processes.
The site is readily accessible, and line power is available. Details of
the ground-based remote sensing and intensive ground sampling are given
in a later section.
2.6. OTHER RESEARCH SITES/FACILITIES
The Small Regional Study
Area contains at least three sites with active cold-season research programs
and facilities that are relevant to the objectives of this experiment.
There may be others as well. Collaborations with these and other programs
is strongly encouraged.
2.6.1. Niwot Ridge Biosphere Reserve and LTER
Niwot Ridge is a broad alpine ridge located in the Colorado Front Range, 20 miles west of Boulder, CO, and 10 miles east of the Fraser MSA. The area was designated in 1975 as an Experimental Ecological Reserve by the Institute of Ecology, and in 1979 as a Biosphere Reserve by UNESCO, the U.S. State Department, and U.S. Forest Service. In 1980 Niwot Ridge was selected by the National Science Foundation as the alpine tundra component of the Long-Term Ecological Research (LTER) program.
Niwot Ridge has also been the site of extensive atmospheric research. The Mountain Climate Program, initiated in 1952, continues to collect valuable data at five principal sites spanning a 5000 foot altitudinal gradient. The National Oceanic and Atmospheric Administration has sampled atmospheric gases from Niwot Ridge since 1968. The carbon dioxide record is the third longest in the world, and is the only long-term record from a continental site.
Many facilities exist to support research on Niwot Ridge, including an alpine laboratory with line power and fiber optic telecommunications capabilities, two subnivean laboratories with complete micrometeorological instrumentation, snowmelt lysimeters, and a dense network of continuously monitored soil and snow temperature profiles. The University of Colorado Mountain Research Station (MRS) is an interdisciplinary research facility managed by the University's Institute of Arctic and Alpine Research (INSTAAR), and is located at the base of Niwot Ridge. An unimproved road provides access to the ridge from the MRS. Facilities at the MRS are available to support research activities on Niwot Ridge, including laboratories, lodging, a dining hall, bathhouse and laundry, and meeting rooms. The MRS has two snow cats that can go to Niwot Ridge when snow conditions permit. Further information about the Niwot Ridge Biosphere Reserve and LTER is available at:
The Loch Vale Watershed is located in Rocky Mountain National Park, about 30 miles southeast of the North Park MSA. Long-term ecological research and monitoring since 1982 addresses watershed-scale ecosystem processes, particularly as they respond to atmospheric deposition and climate variability. Monitoring of meteorological, hydrologic, and water quality parameters enable use of long-term trends to distinguish natural from human-caused disturbances. Research into snow distribution, hydrologic flow-paths, vegetation responses to N deposition, isotopic transformations of N by forest and soil processes, trace metals, and aquatic ecological responses to disturbance enable us to understand processes that influence high elevation ecosystems. The Loch Vale Watershed Long-term Ecological Research and Monitoring Project is operated by the Biological Research Division of the U.S. Geological Survey. Research objectives of the project include:
http://www.nrel.colostate.edu/projects/lvws
The Storm Peak Laboratory (SPL) is located at 3220-m elevation on the top of Storm Peak, within the Rabbit Ears MSA near Steamboat Springs. Located on a peak with limited upwind vegetation or topography to create local turbulence under normal airflow conditions, SPL is ideally situated for in-cloud measurements [Hindman et al., 1994]. This exposure also frequently allows clear-air physical and chemical measurements of the free troposphere (at approximately the 700 mb level) uncontaminated by the local boundary layer [Borys et al., 1988]. The Storm Peak Laboratory is well suited to any number of research programs related to high elevation environmental conditions, such as mountain meteorology and snow pack chemistry. The site is suitable for instrument testing in harsh environmental conditions such as high winds and rime icing. Continuous or intermittent monitoring of free-troposphere aerosol
and gases is possible. The
siting of lidar and other remote sensing systems for free atmosphere and
cloud probing, studies of mountain-valley air circulations, research on
turbulence and wave generation in the vicinity of orographic barriers,
characterization of ozone variability from stratospheric and boundary layer
sources, providing ground truth for mesoscale modeling in irregular terrain,
and verification of radiative parameters derived from satellite observations
are all examples of studies that are feasible. Various research and support
facilities are available. Further information about the Storm Peak Laboratory
is available at:
http://www.dri.edu/Projects/SPL/