H23F-1030 INVITED
The Development of Terrestrial Water Cycle Applications for SMAP Soil Moisture Data Products
Soil moisture storage sits at the locus of the terrestrial water cycle and governs the relative partitioning of precipitation into various land surface flux components. Consequently, improved observational constraint of soil moisture variations should improve our ability to globally monitor the terrestrial water cycle. However, to date, most evidence for such enhancement has been based on synthetic studies and not actual data. The maturity of existing soil moisture datasets (from e.g. the NASA/JAXA AMRS-E and TMI satellite sensors) provides an opportunity to better describe this potential prior to the anticipated launch of the NASA SMAP mission. Using existing remotely-sensed soil moisture datasets, the presentation will demonstrate the potential for improving satellite-based rainfall accumulation products over land and describe a novel data assimilation strategy for leveraging improved rainfall products to enhance global runoff modeling. Despite well-known shortcomings in existing satellite soil moisture data sets (e.g. limited accuracy over vegetation and shallow vertical measurement depths), these strategies lead to measurable improvements in rainfall and runoff estimates over a large fraction of global continental areas. Realized benefits are most profound in lightly-vegetated areas amenable to satellite estimation of surface soil moisture and data-poor land areas lacking adequate ground-based instrumentation. The ability to enhance precipitation also allows for dual data assimilation strategies in which remotely-sensed soil moisture is used to simultaneously correct both the representation of antecedent soil moisture in a hydrologic model and the precipitation forcing applied to the model. Prospects for applying such a dual assimilation approach to data poor areas of Africa will be examined as will potential enhancements associated with the improved accuracy and resolution of SMAP soil moisture products relative to existing datasets.
H23F-1031 INVITED
Soil Moisture Active Passive Validation Experiment 2008 (SMAPVEX08)
The Soil Moisture Active Passive Mission (SMAP) is currently addressing issues related to the development and selection of soil moisture retrieval algorithms. Several forums have identified a number of specific questions that require supporting field experiments. Addressing these issues as soon as possible would contribute to mission definition and algorithms design and selection. Specific objectives include an evaluation of how well do new alternative radio frequency interference (RFI) suppression techniques under consideration for SMAP work over RFI contaminated land areas, providing more robust sets of concurrent passive and active L-band observational data including temporal change for algorithm development and validation, understanding the scaling of high resolution synthetic aperture radar (SAR) to lower resolution radar data of SMAP, and a more thorough evaluation of less studied land covers. A series of aircraft-based flights (SMAP Validation Experiment 2008-SMAPVEX08) was conducted on the Eastern Shore of Maryland and Delaware in the fall of 2008. Several aircraft carrying prototypes of the SMAP instrumentation (combined active and passive microwave) were flown concurrent with ground sampling. Flights were designed to provide a comparison of the various instrument calibrations. Preliminary results will be presented.
H23F-1032
Estimation of Soil Moisture with L-band Multi-polarization Radar
During past years, investigations have demonstrated the capability of active microwave instruments in global soil moisture mapping. However, natural variability and the complexity of the vegetation canopy and surface roughness significantly affect the sensitivity of radar backscattering to soil moisture. Backscattering signals from vegetated areas is a function of water content and its spatial distribution as determined by vegetation structure and underlying surface conditions. It is clear that vegetation cover will cause an under-estimation of soil moisture and an over-estimation of surface roughness when we apply the algorithm for bare surface to vegetation covered regions. Due to complexity of natural surface and vegetation structure (unknowns are more than measurements), however, it is quite difficult to develop a quantitative algorithm to estimate soil moisture in vegetated areas. This study investigates the techniques to estimate surface soil moisture of bare and short vegetated surfaces under SMAP radar sensor configuration: L-band (1.26 GHz) multi-polarization (VV, HH, and VH) radar measurements. We first established a model simulated database using a radiative transfer model. For the surface scattering components, it uses the AIEM model with the random rough surface assumption to simulate the wide range of soil moisture and roughness conditions for co-polarized signals and the Oh's semi- empirical model to simulate the cross-polarized signals. For the vegetation scattering and surface-volume scattering components, we simulated with randomly orientated disk and short cylinders with the maximum of optical thickness and single scattering albedo up to 0.4 and 0.2, respectively. Through analyses of the model simulated database, we developed a technique to estimate surface soil moisture. This technique includes two steps. First, it decomposes the total backscattering signals into two components - the surface scattering components (the bare surface backscattering signals attenuated by the overlaying vegetation layer) and the sum of the direct volume scattering components and surface-volume interaction components at different polarizations. The surface scattering component has the maximum sensitivity to soil moisture and is actually needed information. The signal from the sum of the direct volume scattering components and surface-volume interaction components is the noise signal in terms of the estimating surface soil moisture. From the model simulated database, our decomposition technique works quit well in estimation of the surface scattering components with RMSEs of 0.23, 0.21, and 0.54 dB for VV, HH, and VH polarizations, respectively. Then, we use the decomposed surface backscattering signals to estimate the soil moisture and the combined surface roughness and vegetation attenuation correction factors with all three polarizations. We will show the validation of this algorithm with ground soil moisture and the airborne PALS measurements.
H23F-1033
Soil Moisture Observations by SOMS, GCOM-W1 and SMAP in Mongolia
We have been continuously carrying out ground-based monitoring of soil moisture in the study area in the Mongolian plateau since 2000 with making a validation of the AMSR-E soil moisture measurement algorithms, in order to know the real state of soil moisture behaviors in the central-east Asia. The successful monitoring and validations of AMSR-E have been done in the study area. SMOS and GCOM-W1 will be launched in early summer of 2008 and early 2012, respectively. Synergy observations by SMOS, GCOM-W1 and SMAP will be strongly effective in carrying out precise monitoring of soil moisture and thaw in Mongolia and its surrounding countries that are mainly covered with the cold steppe. Furthermore, we will be able to expect mutually the advance level validations of soil moisture algorithms of SMOS, GCOM-W1 and SMAP. This study discussed in advance a possibility, usefulness and significance of soil moisture observations by SMOS, GCOM-W1 and SMAP in Mongolia based on the results of studies of AMSR-E soil moisture algorithm validation in the study area in the Mongolian plateau.
H23F-1034
Potential L-Band Aquarius Radiometer and Scatterometer Soil Moisture Products from an Observing System Simulation Experiment
Research on the applications of remote sensing techniques has demonstrated the capabilities of microwave observations at L-band wavelength for measuring surface soil moisture. A new source of L-band sensor data, provided by instruments board on Aquarius satellite, is introduced and investigated. Planned for launching in 2010, though primarily dedicated to measuring sea surface salinity (SSS), Aquarius has the potential for providing global surface soil moisture measurements. Based on Aquarius frequencies and viewing characteristics, we are working on generation of the potential Aquarius soil moisture retrieval data, and evaluation and optimization of the impacts of the data products using the NASA's Land Information System (LIS), a powerful land modeling and assimilation system. This simulation of the Aquarius soil moisture retrieval process is being accomplished by means of an OSSE approach. Following the similar effort that resulted in the Hydros OSSE study, we investigate the capability of near feature Aquarius L-band passive and active sensor measurements for monitoring earth's soil moisture globally. Aquarius OSSE includes four key components for simulating soil moisture retrievals: a) land surface modeling (nature run) using LIS-based CLM; b) microwave emission and backscattering modeling; c) orbit and sensor modeling and d) soil moisture retrieval modeling. These components simulate major physical processes involved in data acquisition by the Aquarius radiometer and scatterometer measurements and data conversion to geophysical variables such as soil moisture fields. The first two modeling procedures are conducted at finer 0.125° grid spacing, and the rest of two procedures at coarser 1°, approximately 100 km satellite footprint scale. Linear aggregation is applied for the grid conversion. Aquarius OSSE is carried out over the central United States including the Red Arkansas river basin. One-year simulation period between October 2002 and September 2003 was selected to insure a fair test of the retrieval algorithm performance considering seasonal variations of diverse land surface conditions. We will present all kinds of possible soil moisture maps derived from coarse resolution Aquarius data depending on frequencies and incidence angles. We will make series of comparison of the synthetically derived soil moisture estimates against the CLM nature run to characterize the uncertainties due to land surface heterogeneity, instrument error, and parameter estimates. This comparison will provide new insights into the values of potential L-band Aquarius observations.
H23F-1035 INVITED
The SMAP Algorithm Development Testbed: A Simulation Environment for Algorithm Development and Mission Design Studies
We describe the design and capabilities of a software tool that will be used to simulate observations acquired by the Soil Moisture Active and Passive (SMAP) mission. The SMAP Algorithm Development Testbed is designed to provide an assessment of the soil moisture and freeze/thaw measurement capability for SMAP and to evaluate how this capability will be impacted by various science, instrument, and/or mission trades. The Testbed will assist in evaluating the relative merits of candidate microwave models, retrieval algorithms, and ancillary data for meeting the SMAP soil moisture and freeze/thaw science objectives based on a common set of input and processing conditions. The Testbed has the ability to perform realistic end-to-end simulations involving native orbital and instrument sampling, forward microwave emission and radar backscatter modeling, as well as noise and error source modeling due to instrument characteristics and geophysical effects. The Testbed will be modular by design, and will be presented in an intuitive software interface to encourage independent or collaborative investigations. The Testbed includes the following components: Land Surface Model, Orbital Sampling, Forward Microwave Model, Environmental Effects, Instrument Effects, Inverse Model, and Error Analysis. It is expected that the Testbed will help foster a wide range of studies critical to mission success, including evaluation of different microwave models and retrieval algorithms, as well as the impacts of different science, instrument and mission trades on SMAPês soil moisture and freeze/thaw measurement capability. In this talk we outline the general design of the Testbed and illustrate some of its simulation capabilities with a few representative scenario studies.
H23F-1036
Integrating Land and Radiative Transfer Models for the SMAP Soil Moisture Retrieval and Applications
The SMAP mission will provide the brightness temperature, the retrieved soil moisture in the top few centimeters (as Level-3 data), and the soil moisture in the deep layers through data assimilation (as Level-4 data). The application of these products and the soil moisture retrieval are related to land models in one way or another. In this talk, based on our recent work, we will discuss some relevant issues. First, the retrieval of soil moisture in the top few centimeters requires ancillary data and one of them is the surface temperature which will most probably come from operational models (including land models and satellite skin temperature assimilation). Furthermore, for the direct assimilation of the brightness temperature in operational weather and hydrometeorological forecasting, the realistic simulation of skin temperature in land models is also crucial. We have come up with a new formulation to significantly improve the simulation of skin temperature in the NCEP operational model. This, in turn, significantly increases the use of satellite skin temperature data in the operational forecasting. Second, the retrieval of soil moisture in the deep layers through data assimilation depends on the land model itself. We have found that the numerical solution of the soil moisture-based Richards equation is incorrect in the NCAR Community Land Model (CLM) and perhaps in most other land models as well. A solution has also been developed, and this significantly affects the simulation of the vertical distribution and temporal variation of soil moisture. Finally, soil moisture from satellite remote sensing and from different land models differs significantly (even if different land models have similar fluxes). Different ideas have been proposed in recent years regarding how to use soil moisture from satellite retrieval (or from one model) to other land models (e.g., using the soil moisture normalized by its porosity; using the soil moisture distribution mapping between data and model output and between two model outputs). With a dedicated soil moisture satellite mission (SMAP), we probably need to move beyond these approaches, and some preliminary ideas will be presented in this direction. These preliminary results presented in a short talk will not provide any definite answers; instead, they are intended to pinpoint some directions for future activities that will help the science definition of the SMAP mission.
H23F-1037 INVITED
Algorithm Development for the Soil Moisture Active and Passive (SMAP) Mission Using PALS Radar-Radiometer Observations During the SMEX02 and CLASIC07 Field Campaigns
We describe the analysis and interpretation of field campaign datasets acquired by airborne remote sensing instruments and in situ sensors during the 2002 Soil Moisture Experiment (SMEX) and the 2007 Cloud and Land Surface Interaction Campaign (CLASIC) conducted in Iowa and Oklahoma, USA, respectively. The datasets contain radar and radiometer observations acquired using the Passive and Active L-Band System (PALS) airborne microwave sensor as well as in-situ ground measurements of volumetric soil moisture and land surface temperature. Detailed vegetation information was also acquired during the SMEX02 campaign. By co-locating the in situ measurements and PALS observations within appropriate space-time windows, datasets are obtained that allow examination of the range of validity of existing microwave emission and radar backscatter models, and performance tests of different radiometer-, radar-, and combined radiometer-radar soil moisture retrieval algorithms. Since the data sets were acquired over the same regions on successive days, tests of temporal-change based algorithms can also be performed. The analyses presented in this study will enable progress to be made towards improved understanding of retrieval approaches for the Soil Moisture Active and Passive (SMAP) mission.
H23F-1038
Field measurement of average soil moisture content using cosmic-ray neutrons: a ground truthing tool for the SMAP mission
Ground validation of the satellite soil moisture measurements is a difficult task, in part because of the scale mismatch: existing in-situ measurement methods yield soil moisture at a point, while satellites give average values over many square kilometers. A new method based on emission of cosmic-ray fast neutrons from soils fills the gap. Fast neutrons are produced in soil and air by secondary cosmic-ray neutrons. They are moderated by hydrogen atoms present in soil water and then diffuse into the atmosphere, where their intensity is inversely correlated with soil moisture. These neutrons in air above the ground are measured with the cosmic-ray probe to provide integrated soil moisture content over a footprint of ~700 m and a depth up to 60 cm. The cosmic-ray soil moisture probe is also sensitive to hydrogen above the surface (in snow or vegetation), but the two sources of hydrogen (below and above the ground) can be separated by adding a second neutron detector that is sensitive to thermal neutrons. Measurements accurate to 2% take less than four hours. With its own power, data logging and telecommunication devices, the cosmic-ray probe can measure soil moisture at any location and send the data in near-real time to a server. SMAP's resolution of 1- 3 km (in the active mode) would require only 2-18 cosmic-ray probes to fully cover the area of one pixel. We will present the cosmic-ray neutron method, discuss its applications and advantages and disadvantages, and explore how the cosmic-ray method could be used as a validation tool for the SMAP mission.
H23F-1039
Use of GPS receivers as a soil moisture network to complement satellite studies
The global distribution and temporal variations of soil moisture are required to study the water and carbon cycles. A global network of in situ soil moisture stations is needed to complement datasets from satellite sensors, such as SMAP. We show that high-precision GPS receivers can be used to estimate fluctuations in near surface soil moisture. This is possible because GPS receivers gather energy from ground reflections in addition to the direct signal that travels between the GPS satellite and receiving antenna. The characteristics of the reflected signal change as soil moisture, and therefore the dielectric constant of the ground, varies. We present data from a field site near Boulder, CO. GPS-derived estimates of soil moisture from a 300 m2 area closely match soil moisture fluctuations in the top 5 cm of soil measured with conventional sensors, including the rate and amount of drying following numerous precipitation events. Given this sensitivity to soil moisture, some of the more than 5000 permanent and continuously operating GPS receivers that exist worldwide could be used to provide near-real time estimates of soil moisture for hydrology, climate, and ecology studies. Like the planned SMOS and SMAP missions, the GPS signals are L-band. Thus, GPS receivers are an optimal in situ data source to combine with future satellite measurements.
H23F-1040
Future Monitoring of Northern Carbon Cycle Dynamics from SMAP
A major goal of the Soil Moisture Active Passive (SMAP) mission as directed by the recent NRC Decadal
Survey is to reduce uncertainty regarding the boreal carbon sink for atmospheric CO2. The planned
freeze/thaw (F/T) state measurement from SMAP will provide a surrogate measure of seasonal frozen and
non-frozen conditions that define the potential growing season for northern ecosystems. The F/T variable
also provides a measure of cold temperature constraints to plant growth and sequestration of atmospheric
CO2. The SMAP mission will provide F/T and soil moisture information with much improved spatial
resolution and sensitivity over current satellite microwave remote sensing observations, and will quantify the
primary environmental (temperature and moisture) controls on land-atmosphere CO2 exchange. New
algorithms are being developed to combine future SMAP soil temperature and moisture retrievals with
vegetation gross primary production (GPP) information from optical/IR sensors such as MODIS to estimate
surface soil organic carbon stocks, ecosystem respiration and net ecosystem exchange (NEE) of CO2
with the atmosphere. Initial testing of these algorithms using AMSR-E microwave remote sensing inputs
indicate RMSE accuracies within the uncertainty of tower CO2 flux measurements, while model
sensitivity studies predict >2-fold accuracy improvement from SMAP over AMSR-E. This information will
provide improved mapping and prediction of northern CO2 source-sink activity, component fluxes (GPP,
ecosystem respiration) and associated environmental controls on these processes, and a direct path to
reducing uncertainty regarding the boreal carbon sink for atmospheric CO2. Portions of this
research were carried out at the Ames Research Center and Jet Propulsion Laboratory, California Institute of
Technology under contract to the National Aeronautics and Space Administration.
http://smap.jpl.nasa.gov/
H23F-1041
Toward Development of a Comprehensive Frozen Soil Algorithm
Understanding soil freezing and thawing processes and their interactions/feedbacks in Earth system is essential for assessing variations in regional water cycles, ecosystem productivity, and processes that link the water, energy, and carbon cycles. This is one of the two key and integral components of the NASA Soil Moisture Active and Passive (SMAP) mission to be launched in 2012. Frozen ground includes soils affected by short-term freeze-thaw cycles, seasonally frozen ground, and permafrost. The long-term average maximum area extent, usually in December or January, of frozen ground (including the active layer over permafrost) is about 48.1 million km2 or 50.5 percent of the Northern Hemisphere land area and additional 6.3 million km2 or 6.6 percent is classified as intermittently frozen ground. The timing and duration, thickness, and area extent of soil freezing and thawing are changing spatially and temporally. In fact, some results indicate that during the past few decades, thickness of seasonally frozen ground has decreased substantially. The estimated maximum extent of frozen ground has decreased by 7 percent in the Northern Hemisphere from 1901 through 2002, with a decrease in spring of up to 15 percent. Satellite remote sensing data show that the onset of thaw in spring and freeze in autumn advanced five to seven dates in Eurasia from 1988 to 2002, while over North America, the timing of surface thaw and subsequent initiation of the growing season in early spring has advanced by about eight days from 1988 to 2001. Although numerous investigations have been conducted on soil temperatures and soil freeze/thaw studies, they only deal with either site-specific process studies, surface freeze/thaw status at local or regional scales, or soil maximum seasonal freeze/thaw depth at the end of the freeze/thaw season. These studies use different approaches and certainly have different products. There is no single and comprehensive frozen ground data product that has daily time step with global coverage. In this presentation, we will provide an overview of current available frozen soil algorithms and products. These products in part meet the increasing demands in the scientific community on frozen soil data. We will assess limitations of current available frozen soil algorithms. We will further provide a roadmap toward development of a comprehensive frozen soil algorithm using current available and future satellite remote sensing data, ground-based measurements, and numerical modeling. The ultimate comprehensive frozen soil algorithm should be able to produce blended daily soil temperatures at various depths and daily soil freeze/thaw depths at regional and global scales. The ultimate products will significantly enhance our understanding of ecological and hydrological processes at local, regional, and global scales.
H23F-1042
Detecting near-surface soil Freeze/thaw cycle using Passive Microwave Satellite Remote Sensing Data over China
The near-surface soil freeze/thaw status has significant impact on the energy, water, and carbon exchange between the land and the atmosphere, surface runoff, plant growth, and ecosystem as a whole. The daily monitoring of the region-scale and long-time series of the near-surface soil surface freeze/thaw state are becoming possible with the launch of SMMR (1978~1987), SSM/I (1987~present) and AMSR-E (2002~present). Furthermore, the proposed SMAP (Soil Moisture Active and Passive) mission specifically designed for monitoring surface soil moisture and freeze/thaw state. This paper investigates changes in the near-surface soil freeze/thaw status over the past 31 years (from Jul. 1987 to Mar. 2008) over China. First we validate the existing frozen soil algorithm using 37GHz vertical-polarization brightness temperature and spectral gradient (SG) between 37GHz and 18/19GHz brightness temperature. We further consider the effect of dominant land use types (cropland, forest land, grassland and barren land) on the near-surface soil freeze/thaw classification. The sample clusters representing the frozen and thawed status for each land use type were taken according to the ground surface temperature at the time of radiometer over-passing. Then, the mean value and the standard deviation from their means of 37GHz brightness temperature and SG for SMMR and for SSM/I for each cluster were calculated. The thresholds are determined by the median of the overlap region of the mean¡À1.64¡Ástd (represents the 90% percent of the samples in the cluster) of the frozen and thawed clusters for each land use type. The result of the surface freeze/thaw status classification combined with the in-situ soil temperature and air temperature observations can be used to analyze the seasonally and inter-annual changes in timing and duration, frequency, and areal extent of the near-surface soil frozen/thawed cycle, and to understand how soil responds to the climate warming. The classification result in the overlap period between SMMR and SSM/I are also discussed.
H23F-1043
Monitoring Seasonal Landscape Freeze/Thaw Processes with SMAP
Major science objectives of SMAP support understanding of processes linking terrestrial water, energy and carbon cycles and quantify net carbon flux. The landscape transition between seasonally frozen and non- frozen conditions occurs each year over more than 50 million km2 of the global biosphere, affecting hydrological and ecological processes and associated trace gas dynamics profoundly. The SMAP suite of data products will include maps of landscape freeze/thaw state derived from L-band radar at 1-3 km spatial resolution with a 2-day refresh rate for the high northern latitudes (i.e. latitudes above 50 degrees north). Satellite active and passive microwave remote sensing can be applied to detect large changes in landscape dielectric properties associated with water transitioning between frozen and non-frozen conditions. In the northern high latitudes, seasonal freeze/thaw transitions associated with this process dominate time-series microwave remote sensing signatures. The SMAP freeze/thaw algorithm employs a temporal change detection scheme to delineate freeze/thaw state changes associated with temporal variations in landscape dielectric properties. Development of the algorithm follows from application of legacy data sets provided by satellite radars, both scatterometers and Synthetic Aperture Radars (SARs), and radiometers. This presentation reviews algorithm development and applications, and associated SMAP science objectives addressed through the derived freeze/thaw data products. This work was performed at the Jet Propulsion Laboratory, California Institute of Technology, and at the University of Montana under contract to the National Aeronautics and Space Administration.