ASSESSING PERMAFROST EXTENT AND CONDITION FROM REMOTELY SENSED IMAGERY
Larry D. Hinzman, International Arctic Research
Center, University of Alaska Fairbanks
Permafrost extent, condition, and processes are specifically identified in the Decadal Survey (NRC, 2007) as an observational need and requirement for understanding climate variability and change. For conditions hidden from direct space view such as permafrost, the NRC has recommended (NRC, 2007, p. 260) that “inferences be drawn from in situ measurements and remotely sensed observations from satellite and suborbital platforms.” To date, however, no strategy or NASA missions specifically address the scientific questions surrounding permafrost degradation (NRC, 2007, Table 9.A.1).
We need to develop the remote sensing techniques to produce maps that delineate areas at risk of permafrost degradation and coastal erosion, and produce vulnerability maps for determining safe building locations and provide information where mitigation efforts should be focused to protect Arctic coastal areas. Both permafrost stability and condition can be sensed remotely using surface expression radar, a depth sounding radar, radiometers, airborne CO2 and CH4 flux measurements, and ground observations. It has been proven that surface expressions associated with permafrost degradation can be detected and used to infer information about ecological and hydrological systems (e.g., Grosse et al., 2006; Jorgenson et al., 2001; Kääb, 2008; Osterkamp et al., 2009; Stow et al., 2003; Yoshikawa and Hinzman, 2003), yet this has not been done over a large scale and gas fluxes associated with thawing permafrost have not been adequately quantified. A NASA research priority should relate surface expressions of degrading permafrost to the ecological, biological, hydrological, and carbon systems over a large spatial area providing extent and rate quantification of degrading permafrost and gas flux. Attention should focus upon different manifestations of permafrost degradation, including (A) impacts on terrestrial ecosystems and trace gas exchange with the atmosphere; (B) thermokarst topography and lake development/shrinkage; and (C) coastal erosion. It should be possible to sense these manifestations of permafrost degradation respectively using (1) multifrequency coherent radar and radiometers, (2) airborne and ground-based CO2 and CH4 flux measurements, (3) LiDAR and hyperspectral imaging, and (4) and year-round ground-based and in situ measurements of permafrost. The size of early thermokarst features is in the range of a few meters (thermokarst pits and sinkholes, thaw slumps, ponds). Growth rates can be several meters (pits and sinkholes) to tens of meters (thaw slumps) per season. Thermokarst lake shore erosion is between 0.25 to more than 7 m/yr depending on lake type, region, and shore configuration.
Surface subsidence is on the order of a few cm to tens of cm per year for very active thermokarst subsidence; tens of centimeters to a few meters for thaw slumps; and tens of centimeters to meters for deep but spatially limited sinkholes (ice wedge degradation). Alterations in soil moisture, soil temperature, and asso-
ciated vegetation changes resulting from permafrost degradation have profound and interacting effects on fluxes of carbon and energy (Vourlitis et al., 1993). Both vegetation composition and structure change with permafrost degradation due to direct alteration of the soil hydrological and thermal regime in addition to secondary changes in soil nutrients (Christensen et al., 2004; Jorgenson et al., 2001; Stow et al., 2004). Changes in vegetation will affect the rate and amount of above- and below-ground new carbon storage as well as the surface energy balance through changes in albedo, permafrost insulation, and evapotranspiration, which in turn feed back into the soil hydrological and thermal regime. Decomposition of soil organic matter (and its form of carbon release as CO2 or CH4) will adjust according to the direct relaxation or enhancement of physiological constraints, the size of the unfrozen organic matter pool, and feedbacks to these factors as well as changes in soil organic matter caused by vegetation.
THE ALASKA SATELLITE FACILITY (ASF): PROVIDING REMOTE SENSING DATA IN SUPPORT OF ARCTIC RESEARCH
Don Atwood, Michigan Tech Research Institute
Rapid ecological change suggests that the Arctic may be a bellwether for the impact of global warming upon more temperate parts of the Earth. This suggests the need for a deeper understanding of baseline processes, environmental drivers, and ecological responses in the high latitudes. Due to the Arctic’s vast size, inhospitable conditions, and poor infrastructure, remote sensing will necessarily play an important role in understanding its evolution. The Alaska Satellite Facility (ASF) of the University of Alaska Fairbanks (UAF) is well positioned to support this research. ASF has operated since 1991 as a NASA ground station and archive of satellite data products, with particular focus on synthetic aperture radar (SAR) sensors such as Seasat, JERS-1, ERS-1, ERS-2, RADARSAT-1, and ALOS PALSAR. More recently ASF has become the archive for JPL’s Uninhabited Aerial Vehicle SAR (UAVSAR) and Airborne Microwave Observatory of Subcanopy and Subsurface (AirMOSS) data, and will provide data for the upcoming Soil Moisture Active Passive (SMAP) mission. To date, ASF has established an archive of approximately 2 PB of satellite imagery; most of which covers Alaska, western Canada, the Bering Sea, and the Arctic Ocean.
The goal of this talk is to introduce Arctic researchers to the wealth of data and tools that are available through ASF. For example, long-term historical data sets can be useful for visualizing the nature and evolution of North Slope lakes. In a recent project, ERS-1 and -2 SAR data were used to characterize the bathymetry for all North Slope lakes. More recent polarimetric data from ALOS PALSAR and UAVSAR can provide important microwave scattering data, to assist in understanding land cover/land change, as well as the delineation of wetlands. Beginning in 2015, ASF will begin distributing SMAP data which will be extremely useful for understanding permafrost distribution and dynamics. SMAP products include global freeze/thaw and soil moisture maps (updated every 3 to 10 days), as well as imagery from the onboard SAR and radiometer instruments. With an anticipated mission life of 3 years, SMAP will capture the entire annual hydrological cycle as well as chronicle interannual variations attributable to melting permafrost. The presentation will finish with a brief description of how ASF data can be freely acquired for research purposes.
WHAT LIES BENEATH: AIRBORNE ELECTROMAGNETIC METHODS FOR MAPPING SUBSURFACE PERMAFROST AND BUILDING GEOLOGICAL FRAMEWORKS IN COLD REGIONS
Burke J. Minsley, U.S. Geological Survey, Crustal
Geophysics and Geochemistry Science Center, Denver, CO
Airborne electromagnetic (AEM) methods play a unique role in the remote sensing of permafrost, and related geological and hydrological environments, because of their ability to map the subsurface from depths of a few meters to several hundred meters below ground. In fact, AEM is not often grouped with more traditional remote sensing technologies, but rather is classified with geophysical methods—a somewhat arbitrary and misinformative distinction. AEM is the only available remote sensing tool that helps to bridge the gap between other airborne and satellite technologies
that map surface (or very shallow) features over large areas, and the sparse ground truth information about physical properties at depth from borehole data.
AEM is a decades-old technology borne out of the mineral exploration industry, but has recently seen widespread application in geological and hydrological mapping programs, as well as permafrost and sea-ice thickness studies, that has been facilitated by improvements in instrumentation and processing methods. AEM relies on the physics of electromagnetic induction, as opposed to many other remote sensing modalities that rely on wave propagation, to detect physical properties from the near surface down to several hundred meters below ground. Depth imaging is achieved by acquiring data at different frequencies (several hundred Hz to approximately 100 kHz), where lower frequencies are sensitive to deeper structures. Inductive electromagnetic methods are primarily sensitive to the electrical characteristics of geological materials that, in turn, are a function of properties such as unfrozen water content, lithology, and salinity. A significant challenge in the interpretation of AEM data is to infer the underlying physical properties that result in the mapped distribution of electrical resistivity.
I will introduce the basic instrumentation and methods behind AEM surveying and interpretation, along with examples and ideas of how AEM data can be integrated with other remote sensing products and ground-based measurements for robust, multiscale mapping of permafrost systems. For example, an AEM survey acquired by the USGS in the Yukon Flats of Alaska revealed the subsurface geometry of discontinuous permafrost, and also captured the thermal legacy of the Yukon River lateral migration over the past ~1,000 years, which has been recorded in permafrost. Other AEM surveys acquired by the Alaska Division of Geological and Geophysical Surveys (DGGS) are being interpreted to better constrain permafrost distributions in diverse and complex geological settings. And a recent NSF-supported AEM survey has mapped glacier ice, permafrost, and saline water in portions of Antarctica’s dry valleys. These results illustrate the value of AEM data for developing three-dimensional geological and hydrological frameworks of permafrost environments, and the importance of furthering the use of AEM to complement our permafrost remote sensing toolbox.
HOW TO IMPROVE PERMAFROST MODELS USING REMOTE SENSING
Kevin Schaefer (NSIDC), Lin Liu (Stanford),
Howard Zebker (Stanford), Andrew Parsekian
(Stanford), Elchin Jafarov (NSIDC), Santosh Panda
(UAF), Tingjun Zhang (NSIDC)
Improving the representation of permafrost is a key factor in improving the performance of global climate models. We need both in situ and remote sensing data of permafrost characteristics and processes for model validation and parameterizations. Measurements of permafrost temperature, active layer thickness, and other characteristics provide validation or initial values for model prognostic variables. Measurements of landscape changes due to various permafrost processes provide validation data and inputs to model parameterizations. A model parameterization is an observationally based, statistical representation of the large-scale effects of subgrid processes. Models use parameterizations to represent processes that occur on a physical scale that is much smaller than the model resolution, such as wetland dynamics, hydrology, runoff, erosion, fire, insect infestation, snow dynamics, and thermokarst. All of these parameterizations need improvement, particularly parameterizations of thermokarst processes, which are exceedingly rare in current global models.
We recommend a remote sensing strategy designed to improve model parameterizations of permafrost processes. First, you identify surface properties known to reflect large-scale effects of permafrost processes and modify models to simulate these properties. You use remote sensing and in situ data to measure how these properties change over time and develop statistical relationships between the measurements and the effect of a specific process. The model parameterization is the statistical relationships applied to the simulated property to estimate the bulk effects of a permafrost process. To illustrate the strategy, we show how to use interferometric synthetic aperture radar (InSAR) and optical remote sensing data to develop a parameterization of thermokarst lake expansion.
We also recommend expanding the use of geophysical remote sensing techniques to complement current and planned remote sensing capabilities. Geophysical remote sensing includes InSAR, electro-
magnetic methods, ground-penetrating radar, nuclear magnetic resonance, and related techniques. These and other techniques are widely used on a variety of platforms, including manual deployment, vehicles, aircraft, and satellites. Field measurements taken with such instruments are often lumped with in situ data, but they are, in fact, remote sensing. Geophysical remote sensing often leverages differences in the physical properties of water, ice, and soil, the main ingredients of permafrost. Geophysical techniques have been widely applied to the cryospheric study of glaciers, land ice sheets, and sea ice, but are greatly underutilized in the study of permafrost. We offer a number of potential examples to illustrate the full potential of geophysical remote sensing in understanding key permafrost processes.
SPATIAL AND TEMPORAL RESOLUTION REQUIREMENTS FOR REMOTE OBSERVATION OF PERMAFROST LANDSCAPE DYNAMICS
Permafrost, at regional scales, can be seen as a relatively homogeneous subsurface property—a state of ground temperature defined by areal extent or average vertical thickness. However, at local scales heterogeneity in ground ice content and distribution, soil organic layer thickness, soil stratigraphy, and external factors such as snow and vegetation distribution, micro- and meso-topography, and hydrological framework lead to variability in the response of permafrost landscapes to change. Thus, the scale at which observations are made is important for detecting local disturbances of the ground thermal regime that may lead to thermokarst and other thaw-related landscape features.
Land surface features and processes often allow derivation of information about the local state of permafrost in an area. Identification, mapping, and monitoring of such features/processes with remote sensing can provide access to various subsurface properties and dynamics of near-surface permafrost and the active layer, greatly helping in understanding vulnerabilities and trajectories of change. An important question is what spatial and temporal resolution requirements have to be met by remote sensors to adequately capture local permafrost features and landscape dynamics useful for interpreting the local state and vulnerability of permafrost.
In our presentation we will show examples of permafrost landscape features and processes and discuss what their spatial scales and temporal dynamics are, including thermokarst pond growth, lake and coastal erosion, thaw slump development, peatland collapse, changes in active layer thickness, broad surface subsidence, as well as pingos, ice wedge networks, and small-scale patterned ground. We will differentiate between seasonal versus long-term changes and forward the notion that permafrost change may express both as degradation as well as aggradation.
We (1) will highlight a range of past and current optical remote sensors and their capabilities and limitations in capturing these permafrost landscape dynamics at sufficient spatial and temporal resolution. We will further discuss the application and need for (2) highly accurate and high-resolution digital elevation models, (3) repeat acquisition of elevation and optical remote sensing data sets, and (4) multisensor approaches joining optical and SAR capabilities.
CURRENT STATUS AND FUTURE OF SATELLITE REMOTE SENSING TO BETTER UNDERSTAND ECOLOGICAL IMPACTS TRIGGERED BY CHANGING PERMAFROST
Dara Entekhabi, MIT
The NASA Soil Moisture Active Passive (SMAP) is due to launch in November 2014. The mission provides for global mapping and monitoring of landscape freeze/thaw (FT) status and surface soil moisture conditions. The SMAP Level 2/3 FT product will quantify the predominant frozen or nonfrozen status of the landscape at approximately 3-km resolution and 3-day fidelity. The FT retrievals will be validated to a mean spatial classification accuracy of 80%, sufficient to quantify frozen season constraints to terrestrial water mobility and the potential vegetation growing season
1Geophysical Institute, University of Alaska Fairbanks.
2Alaska Science Center, U.S. Geological Survey, Anchorage.
over northern (≥45°N) land areas. A SMAP Level 4 carbon (L4_C) product uses the FT retrievals and model value-added surface and root zone soil moisture estimates with other ancillary inputs to quantify net ecosystem CO2 exchange (NEE), component carbon fluxes, and surface (<10 cm depth) soil organic carbon (SOC) stocks over all global vegetated land areas. The L4_C product also quantifies underlying environmental controls on these processes, including soil moisture and frozen season constraints to productivity and respiration. The L4_C NEE estimates will be validated to an RMSE requirement of 30 g C m2 yr–1 or 1.6 g C m2 day–1, similar to accuracy levels determined from in situ tower eddy covariance CO2 flux measurements. The L4_C research outputs include SOC, vegetation productivity, ecosystem respiration, and environmental constraint (EC) metrics clarifying FT and soil moisture–related restrictions to estimated carbon fluxes. These products are designed to clarify how ecosystems, especially in boreal regions, respond to climate anomalies and their capacity to reinforce or mitigate global warming.
The SMAP mission provides for global mapping of soil moisture and landscape FT state dynamics with enhanced L-band (1.26/1.41 GHz) active passive microwave sensitivity to surface soil conditions, and approximate 3-day temporal revisit owing to its 1000- km-wide swath. The radar resolution is better than 3 km in the outer 70% of the swath and away from the satellite track at nadir. The SMAP satellite is in a polar orbit which results in considerable overlap of swaths in northern latitudes. By combining the far outer edge of overlapping swaths each day, it is possible to construct a daily L-band radar mapping of northern latitudes at 1 km resolution. The measurements are valuable to monitoring the land and sea cryosphere regardless of clouds, weather, and solar illumination. We propose a community effort to produce an all-Alaska daily 1-km-resolution L-band radar backscatter cross-section product based on Level 1 SMAP files at the Alaska Satellite Facility (NASA-designated Distributed Active Archive Center for SMAP radar data).
AIRBORNE REMOTE SENSING CAPABILITIES TO UNDERSTAND ECOLOGICAL IMPACTS TRIGGERED BY CHANGING PERMAFROST
Charles Miller, Jet Propulsion Laboratory, California
Institute of Technology
The presentation includes a survey of airborne instruments and techniques for tracking high-latitude ecology and atmospheric carbon dioxide and methane. There is also a brief survey of existing airborne assets for permafrost characterization, including both radar and electromagnetic (EM) methods.
There are two significant ecological impacts of permafrost change. One is a “greening Arctic,” which results in a change in species and range of vegetation cover, and an increase in carbon uptake. Second is an increase in carbon dioxide and methane emissions from mobilized ancient carbon. Carbon dioxide/methane fractioning depends critically on changes in hydrology. The future carbon budget of the northern high latitudes depends on the (im)balance that a changing permafrost imposes.
Airborne remote sensing also offers the potential for multisensor observations that can bring unique insights into the ecosystem processes and properties. Asner et al. (2007) pioneered the fusion of high-fidelity visible/VIS-SWIR hyperspectral imaging spectrometer data with scanning, waveform light detection and ranging (wLiDAR) data, along with an integrated navigation and data processing approach. This is a quantum leap beyond BOREAS ecosystem remote sensing. It retrieves information on vegetation canopy structure, vegetation biochemistry, vegetation biophysical properties, and the ecosystem response. It will be available on a regular basis in Alaska via NEON (National Ecological Observatory Network) beginning around 2014.
Solar induced chlorophyll fluorescence (SIF) can be measured through high-spectral-resolution remote sensing in the 690-770 nm region. SIF is directly related to photosynthetic activity, although the exact functional relationship of SIF to GPP (Gross Primary Production) is currently debated. SIF is measured by airborne sensors (FLEX Simulator, CARVE FTS, MAMAP) and satellite instruments (GOSAT, OCO-2, OCO-3, FLEX [proposed], CarbonSat [proposed]).
It is also possible to measure the total column carbon dioxide and methane using high-spectral-resolution remote sensing in the 1650-2400 nm region. Airborne sensors with this capability include CARVE FTS and MAMAP. Satellites that can measure carbon dioxide include GOSAT, OCO-2, OCO-3, and CarbonSat (proposed). Satellites that can measure methane include SCIAMACHY (no longer available), GOSAT, CarbonSat (proposed), and the Sentinel 5 precursor. It is important to note that NASA currently has no plans for a space-based mission to measure methane over the high latitudes.
The presentation also includes several examples of airborne methods that can be used for permafrost characterization: (1) the AirMOSS Flight System; (2) the Boreal Ecosystem Research Monitoring Sites (BERMS); (3) mapping the average seasonal subsidence between 1992 and 2000 near Prudhoe Bay, Alaska, by utilizing a time series of 14 interferograms from the ERS satellite; (4) mapping the average active layer thickness (ALT) between 1992 and 2000 near Prudhoe Bay by converting the seasonal subsidence to melted water and assuming a vertical water distribution; (5) resistivity cross sections in the Yukon-Flats Region; and (6) three-dimensional mapping with HEM (helicopter electromagnetic).
A GEOBOTANICAL PERSPECTIVE: MONITORING ARCTIC PERMAFROST AND ECOSYSTEM CHANGE USING REMOTE SENSING, GIS, AND PLOT-BASED STUDIES
D.A. Walker, Institute of Arctic Biology, University of
Alaska Fairbanks, USA
Integrated mapping approaches for the Arctic have been evolving. Since beginning in 1969, the Alaska Geobotany Center (AGC) has made vegetation maps using traditional photo-interpretive methods, satellite sensors, and plot-based interdisciplinary research along environmental gradients. These maps have a myriad of applications to permafrost and global-change research. In this talk I discuss four points that I see as essential for developing a comprehensive interdisciplinary approach to use remote sensing to examine Arctic change:
Spatial hierarchy of databases: Databases are needed for answering questions at plot to planet scales. This necessarily requires consistent approaches for visualizing and coloring the maps so that they make intuitive sense across scales. Recent availability of very-high-resolution satellite imagery promises to revolutionize interpretation of changes to permafrost patterned-ground landscapes.
Circumpolar databases: A circumpolar examination of permafrost and environmental change requires pan-Arctic spatial databases. Such databases require a high level of synthesis and international coordination.
Long-term data sets: Time series of ground observations need to complement time series of remote sensing images and detailed mapping.
Integrated mapping studies: Mapping should integrate as much geoecological information from different disciplines as possible into single databases along with historical natural geoecological changes and anthropogenic changes. Examples include the Integrated Terrain Unit Mapping approach developed by ESRI Inc. and the “integrated geoecological and historical change maps” (IGHCM) that were developed to simultaneously examining historical changes caused by dynamic permafrost landscapes and those caused by expanding networks of oil field infrastructure.
The AGC and the Geographic Information Network of Alaska (GINA) are in the process of developing an Arctic Alaska Geoecological Atlas for NASA’s planned Arctic Boreal Vulnerability Experiment (ABoVE) that will draw on the principles discussed above (http://above.nasa.gov).
THE CONTRIBUTION OF SPACEBORNE SYNTHETIC APERTURE RADAR SENSORS TO PERMAFROST RESEARCH
Franz J. Meyer, Associate Professor, University of Alaska Fairbanks, AK
Paul A. Rosen, Jet Propulsion Laboratory, Pasadena, CA
In the recent decade, data synthetic aperture radar (SAR) sensors have been shown to have great potential for observing the Arctic. This is in large part due to two advantageous characteristics of SAR data: (1) As an active sensor, SAR systems can observe the ground independent of weather and illumination conditions
and are as such the only systems that can reliably provide 24/7 observations and (2) in addition to image information, SAR phase observations can provide measurements of surface dynamics, which can be used to indicate surface change and measure centimeter-scale surface deformation. Owing to these benefits, SAR data have the potential to provide information on two major processes in permafrost regions that are relevant for understanding climate change impacts in northern high-latitude environments: short-term, seasonal dynamics of the active layer located above permafrost, and long-term multiannual changes in permafrost extent.
We will show in theory and examples that current and future SAR missions can provide information on several key parameters for seasonal active layer (freeze/thaw, subsidence and heave, deep soil moisture) and long-term permafrost dynamics (subsidence, lateral movements). We will furthermore particularly highlight the planned capabilities of an upcoming proposed NASA L-band SAR mission that can provide permafrost-related information at high spatial and temporal resolution and at accuracy levels that may lead to a substantial improvement of our understanding of panarctic active layer dynamics and permafrost thaw. We will summarize the mission’s predicted measurement characteristics by highlighting its high temporal and spatial sampling, its global observation strategy, and its predicted performance in measuring surface dynamics. Based on the system’s proposed measurement capabilities, we claim that this future L-band mission will allow for a spatially explicit assessment of regional to global impacts of permafrost dynamics on hydrology, carbon cycling, and northern ecosystem character and functioning.
MICROWAVE REMOTE SENSING UTILITY FOR DOCUMENTING ENVIRONMENTAL CHANGE IN PERMAFROST LANDSCAPES
John S. Kimball, Flathead Lake Biological Station,
Division of Biological Sciences, The University of Montana
This presentation highlights the potential utility of satellite active and passive microwave remote sensing for regional monitoring of physical attributes related to soil active layer dynamics and permafrost in the boreal/arctic. Relative strengths and limitations of current global satellite records and the potential utility of upcoming NASA Earth missions are discussed. Potential research gaps are identified and recommendations for improving the relevance of these observations for permafrost landscapes are presented.
Satellite active and passive microwave remote sensing at lower frequencies (~≤37 GHz) has strong utility for mapping and monitoring of physical land parameters relevant to soil active layer dynamics in permafrost landscapes. Satellite passive microwave sensors detect natural microwave emissions of the land surface, while the associated brightness temperature (Tb) retrievals are strongly sensitive to surface moisture and temperature through their effect on surface dielectric properties. Because only a small portion of Earth’s energy emissions are at lower microwave frequencies, the Tb retrievals have generally coarse (~12-60 km) spatial resolution to enhance the sensor signal-to-noise ratio. In contrast, active microwave sensors (radars) provide their own land surface illumination source, enabling finer-spatial-scale retrievals with a larger signal-to-noise ratio. The sensitivity to soil attributes is strongly dependent on microwave frequency and land surface conditions. Lower frequencies (e.g., L/P band) have generally greater potential soil active layer sensitivity, while the relative depth of direct soil sensitivity is inversely proportional to the moisture content in soil and overlying snow and vegetation layers. The relative insensitivity of microwaves to solar illumination and atmosphere aerosols, including clouds and smoke, and the converging orbital swaths of polar orbiting sensors enable daily or better temporal fidelity over northern (≥50°N) land areas from operational global satellites, while finer-scale synthetic aperture radar (SAR) sensors have more limited spatial and temporal coverage. These attributes have been exploited for monitoring a range of physical parameters, including surface soil moisture and temperature, landscape freeze/thaw dynamics, open water inundation, snow cover, vegetation biomass, and terrain structure. Similar satellite microwave retrievals from overlapping operational sensor records have also enabled the development of relatively long-term records documenting recent (up to 30+ year) environmental changes with relatively high precision.
New global satellite missions are coming online
that enable new observations of biophysical attributes in permafrost landscapes. The NASA SMAP (Soil Moisture Active Passive) mission has a projected launch in mid-2014 and will provide enhanced L-band (1.26/1.4 GHz) sensitivity to surface soil moisture and landscape freeze/thaw dynamics with regular global monitoring at moderate (1-9 km) spatial resolution and 1-3 day temporal repeat. Model-enhanced (Level 4) products are also planned that will provide daily estimates of soil profile (≤1 m depth) moisture and thermal conditions, surface soil organic carbon (SOC) stocks, and terrestrial carbon (CO2) fluxes. However, a number of limitations remain for applying existing and planned satellite sensor records for monitoring environmental change in permafrost landscapes. These limitations include a lack of finer resolution (e.g., 10-100 m) monitoring approaching the scale of landscape variability in permafrost attributes. Other limitations include loss of direct sensitivity to soil attributes under higher vegetation biomass and potentially complex processing required for extracting meaningful land parameter information from lower order microwave retrievals. Various methods have been developed for regional downscaling of satellite observations that may enhance utility of these data for permafrost. These techniques include spatial resolution enhancement techniques applied to overlapping Tb and radar backscatter orbital swath data, and empirical modeling and data fusion techniques using synergistic multiscale and multisensor remote sensing, and other ancillary data for estimating finer-spatial-scale attributes. While these techniques have been widely used for other areas, there is still a general paucity of regional applications of these techniques in permafrost landscapes.
New airborne assets have become available that can be used to inform algorithm development and regional downscaling efforts. These assets include the NASA AirMOSS and UAVSAR sensors, which provide for finer-scale L/P-band SAR retrievals. The JAXA JERS-1 SAR and ALOS PALSAR sensor records provide similar fine-scale (~10 to 100 m resolution) L-band satellite SAR data, which have been used to investigate sub-grid-scale landscape freeze/thaw heterogeneity. These initial studies indicate potential utility for quantifying scaling behavior in similar, coarser-scale satellite retrievals that can be used to inform regional downscaling efforts. Similar L-band SAR data will be provided by ALOS-2, which has a projected launch in late 2013. However, user access to these data may be severely constrained by limited access and cost-per-scene data use restrictions, and may require significant NASA investment in potential data buys or data use agreements to secure unrestricted access to these data.
Recent studies using AirSAR and theoretical microwave radiative transfer modeling indicate potential utility for combined L/P-band SAR remote sensing for direct retrievals of soil active layer development in lower biomass (e.g., tundra) areas, while reduced soil sensitivity under higher biomass (e.g., boreal forest) conditions may constrain direct soil retrievals in the boreal zone. However, these constraints may be mitigated by using additional vegetation biomass structure information provided by LiDAR and optical sensors. New investments in coordinated satellite and airborne remote sensing and detailed ground network measurements are needed for further algorithm development and validation (Cal/Val) efforts to fully develop and demonstrate these capabilities.
Several near-term (next 3-5 years) boreal-Arctic field campaigns are under development, including limited campaigns supporting post-launch sensor and product Cal/Val activities for SMAP, OCO-2 (Orbiting Carbon Observatory), and a more extensive NASA-led Arctic Boreal Vulnerability Experiment (ABoVE). These campaigns will involve coordinated satellite, airborne, and field-based measurements, with potential focus on physical parameters directly relevant to permafrost attributes and processes. These activities provide opportunities for testing sensors and developing retrieval algorithms specific to permafrost landscapes. Finally, a number of agency and international efforts are under way or in the planning stages that focus on improving monitoring capabilities and understanding of environmental change in permafrost landscapes. Collaborative partnerships among these agencies and efforts should be encouraged for scoping and communicating joint needs, developing new satellite missions and field campaigns, and the free exchange of data.
POTENTIAL OF HYPERSPECTRAL REMOTE SENSING FOR MAPPING PERMAFROST FEATURES AND ASSOCIATED BIOPHYSICAL VARIABLES
Anupma Prakash, Jordi Cristóbal, Christian
Haselwimmer, and Don Hampton, Geophysical Institute,
University of Alaska, Fairbanks
Hyperspectral remote sensing, also known as imaging spectroscopy, has potential to make a significant contribution in permafrost research, by providing a tool to map associated biophysical variables with unprecedented detail. A large body of literature exists that documents the success of imaging spectroscopy in identifying and mapping plant species and plant functional types. However, literature on the direct use of imaging spectroscopy for mapping permafrost features or linking permafrost distribution with biophysical variables identified and mapped by imaging spectroscopy is at best limited.
NASA’s planned Hyperspectral Infrared Imager (HyspIRI) mission is designed to quantitatively study the Earth’s terrestrial biosphere, identify vegetation species and functional types, and provide benchmark mapping against which future changes can be assessed. With two imaging spectrometers, one in the 380 to 2500 nm visible shortwave infrared (VSWIR) region, and the other in the 3 to 12 mm thermal infrared (TIR) region, providing 60 m spatial resolution data and a near global coverage every 19 days (for VSWIR)/5 days (for TIR), the instrument will support a broad spectrum of carbon and water cycle and ecosystem studies. Permafrost research will particularly benefit from HyspIRI’s VSWIR instrument that will provide a means for superior identification and classification of Arctic and sub-Arctic vegetation. The potential for better characterizing the mosses and shrubs associated with permafrost landscape will be a distinct advantage. The TIR instrument will allow us to better map energy fluxes and other related biophysical variables (such as soil moisture or air temperature) at medium spatial resolution.
At the University of Alaska Fairbanks (UAF) we have invested in two areas that will support permafrost research: (i) Through a recent grant from the National Science Foundation Major Research Instrumentation (MRI) program, UAF is in the initial stages of developing an in-state capability (The University of Alaska Fairbanks Hyperspectral Imaging Laboratory—UAF HyLab) for airborne and ground-based imaging spectroscopy based around the acquisition of commercial HySpex visible and shortwave infrared (0.4-2.5 µm) hyperspectral systems. The capability for routine acquisition of new imaging spectroscopy data sets over Alaskan study sites will provide a tremendous boost to the permafrost remote sensing, and for studying ecosystem composition and change. (ii) Through a NASA EPSCoR grant, UAF has established two field sites, one in the interior Alaska boreal forest setting and another in the deciduous forest setting, which include a suite of ground-based instrumentations collecting data on surface energy flux, ground heat flux, and other essential climate variables. These field sites, that can serve as calibration and validation (CalVal) sites for satellite missions, are being used to scale observations from field scale to satellite scale.
This page intentionally left blank.