P.I. Ted Engman, Hydrological Sciences Branch, NASA/GSFC

Co.I. Peggy O'Neill, Hydrological Sciences Branch, NASA/GSFC

Co.I. Eric Wood, Water Resources Program, Princeton University

Contributors:

Valentine Pauwels, Water Resources Program, Princeton University

Ann Hsu, SSAI, Hydrological Sciences Branch, NASA/GSFC

Tom Jackson, Hydrology Lab, USDA-ARS

Visiting Scientists:

J.C. Shi, University of California Santa Barbara

Corinna Prietzsch, ZALF, Muncheberg, Germany

Peter van der Keur, University of Copenhagen, Denmark

Kazuhiku Fukami, Public Works Research Institute, Japan

INTRODUCTION

The Shuttle Imaging Radar-C and X-Band Synthetic Aperture Radar (SIR-C/X-SAR) is a cooperative space shuttle experiment between the National Aeronautics and Space Administration (NASA), the German Space Agency (DARA), and the Italian Space Agency (ASI). The experiment is the next evolutionary step in NASA's Spaceborne Imaging Radar (SIR) program that began with the Seasat Synthetic Aperture Radar (SAR) in 1978, and continued with SIR-A in 1981 and SIR-B in 1984. It also represents a continuation of Germany's imaging radar program which started with the Microwave Remote Sensing Experiment (MRSE) flown aboard the Shuttle on the first SPACELAB mission in 1983. The SIR-C/X-SAR Mission benefits from synergism with the Magellan Mission to Venus, other international spaceborne radar programs, and prototype aircraft sensors such as the JPL Airborne SAR (AIRSAR) and the German Aerospace Establishment (DLR) E-SAR.

The SIR-C/X-SAR mission extends the capability of an aircraft campaign by providing regional-scale microwave data on a rapid temporal scale. The mission design also enables areas to be imaged at multiple aspect and incidence angles, which is important for studying many land and ocean processes. The extensive surface measurement campaigns provided critical data to be used in development of algorithms needed to produce key geophysical products for assessing global change issues. By having multiple flights, insights on seasonal variations for the key science issues was provided. Such validation and algorithm development studies are critical for developing future mission concepts.

OBJECTIVES

The objectives of this research are to:

1. To determine and compare soil moisture patterns within one or more humid watersheds using SAR data, ground based measurements, and hydrologic modeling.

2. To use radar data as a basis for scaling up from small scale, near-point process models to larger scale water balance models necessary to define and quantify the land phase of GCMs.

3. To use radar data to characterize the hydrologic regime within a catchment and to identify the runoff producing characteristics of humid zone watersheds.

STUDY AREAS

Little Washita Watershed. The Little Washita River watershed is a 611 sq km tributary to the Washita River near Chickasha in southwest Oklahoma. The watershed is in the southern part of the Great Plains of the United States. The climate is classified as moist and subhumid with an average annual precipitation of 76 cm. Summers are typically hot and relatively dry, and winters are temperate and dry. Most of the annual rainfall and potential for floods occurs in the spring and fall. The Little Washita watershed is an intensively instrumented research basin operated and maintained by the Agricultural Research Service, U. S. Department of Agriculture (ARS USDA), and was designated as a hydrology supersite for the SIR-C missions.

Mahantango Creek Watershed. The Mahantango Creek watershed served as a backup supersite for hydrology. The Mahantango is in the ridge and valley region of eastern Pennsylvania and is another experimental watershed run by the ARS USDA. The climate is cool and humid with approximately 100 cm of rain evenly distributed throughout the year. The focus of the SIR-C data collection activities was the intensively instrumented E-38 sub-basin (7 sq km) of the 100 sq km upper Mahantango Creek basin above Klingerstown, PA.

DATA COLLECTION

Two intensive field measurement campaigns were conducted in the Little Washita in 1994 coincident with the two SRL missions which took place from April 11 to 17 and from October 2 to 6. The second mission replaced a last second aborted mission in August.. Tables 1 and 2 list the shuttle data takes during these two missions. During both missions, as well as the week of the aborted August shuttle flight, ground and aircraft data were collected at both the primary and secondary supersites. The types and quantity of data collected at the two sites varied from mission to mission but included extensive ground truth data (soil moisture, roughness, vegetation, etc.), flux stations, radiosondes, weather stations, stream flow, and aircraft measurements (C-130 ESTAR, NS001, TIMS, JPL AIRSAR, USDA vis-nir and thermal).

During the SRL-1 field experiment (April 6-16, 1994), field conditions were initially dry, became wet as a result of a heavy rainstorm on the afternoon of 4/10/94 and morning of 4/11/94, and then gradually dried down throughout the remainder of the experimental period (a change in volumetric soil moisture of 15-20%). In contrast, during the SRL-2 experiment (October 1-6, 1994), most of the watershed remained dry except for localized rainfall in the northwest part of the watershed on the evening of 10/3/94 and in the northern part of the watershed on the evening of 10/4/94.

Conditions at Mahantango in April were very wet with little change in soil moisture between the two data takes. During October, conditions were quite a bit better. The area was initially dry for the first data take, followed by rain and a dry down for the next standard data take and the consecutive days of the interferometric data takes.

Table 1. April mission data takes for the Chickasha, Oklahoma supersite.

Date	Orbit	DT	CPA/Time	Flight direction	Look direction	Look Angle (deg)	Incidence Angle (deg)	Mode*
4/11	34	34	2/1:03:31	A	S	25.46	26.52	16X
4/12	50	50.1	3/0:44:37	A	S	37.75	39.32	16X
4/12	55	55.4	3/8:30:49	D	S	56.79	60.05	11X
4/13	66	66.1	4/0:25:24	A	S	46.25	48.44	16X
4/13	71	71.3	4/8:11:31	D	S	52.35	55.09	11X
4/14	82	82.1	5/0:05:54	A	S	52.16	54.81	11X
4/14	87	87.51	5/7:51:55	D	S	47.06	49.32	11X
4/15	98	98.11	5/23:46:05	A	S	56.34	59.48	11X
4/15	103	103.3	6/7:31:59	D	S	40.86	42.69	16X
4/16	119	119.2	7/7:11:44	D	S	33.84	35.27	16X
4/17	135	135.3	8/6:51:09	D	S	26.38	27.47	16X

*MODE 16X: LHH, LHV, LVV, LVH, CHH, CVH, CHV, CVV, XVV

MODE 11X: LHH, LHV, CHH, CHV, XVV

Table 2. October mission data takes for the Chickasha, Oklahoma supersite.

Date	Orbit	DT	CPA/Time	Flight direction	Look direction	Look Angle (deg)	Incidence Angle (deg)	Mode*
10/2	34	34	2/1:03:31	A	S	25.46	26.52	16X
10/3	50	50.1	3/0:44:37	A	S	37.75	39.32	16X
10/3	55	55.4	3/8:30:49	D	S	56.79	60.05	11X
10/4	66	66.1	4/0:25:24	A	S	46.25	48.44	16X
10/4	71	71.3	4/8:11:31	D	S	52.35	55.09	11X
10/5	82	82.1	5/0:05:54	A	S	52.16	54.81	11X
10/5	87	87.51	5/7:51:55	D	S	47.06	49.32	11X
10/6	98	98.11	5/23:46:05	A	S	56.34	59.48	11X
10/6	103	103.3	6/7:31:59	D	S	40.86	42.69	16X

RESEARCH RESULTS

CD ROM. A collection of ground based, aircraft, and satellite data sets acquired as part of the SIR-C/X-SAR hydrology experiment during April and October, 1994 in the Little Washita watershed near Chickasha, Oklahoma, has been prepared for distribution to the scientific community. The CD ROM includes a description of the ground test sites, site location maps, ground photos of selected sites, average soil moisture for each site, aircraft false color composite images from 4/6/94, Landsat Thematic Mapper data from 4/12/94, ERS-1 satellite radar data from 4/15/94, SIR-C/X-SAR data from April, 1994 (dB), SIR-C data from October, 1994 (dB), SIR-C contrast stretched images, and L and C band backscatter from a truck radar system deployed to Chickasha in support of the April mission.

Truck radar experiments. To complement the Shuttle radar data and provide an accurate means of comparison of microwave data acquired at different spatial scales from different platforms, NASA/GSFC's three-frequency truck-mounted radar was deployed to the hydrology supersite near Chickasha during SRL-1. Designed and operated in conjunction with George Washington University, the truck radar consists of L, C, and X band radars (1.6, 4.75, and 10 GHz) which approximate the L, C, and X band radars on SIR-C and the L and C radars on NASA/JPL's airborne AIRSAR system. On each of eight days of actual data collection the truck radar took measurements at 3 frequencies (L, C, X), 4 polarizations (HH, HV, VV, VH), and 3 angles (30, 40, 50 degrees) for 4 test fields which each represented a typical surface cover found in the watershed (pasture, alfalfa, new corn/bare, winter wheat). Calibrated field averages of truck radar backscatter for each of these test fields have been compiled for L and C band and are available for distribution; most of the X band data were considered to be marginal due to noise floor problems. Comparisons of truck radar, SIR-C, and AIRSAR L band data indicate that for certain fields the three sensors were responding almost identically to the soil moisture dry down (agreement generally to within 1 dB). In other fields, there might be agreement in some channels but not in others (like vs. cross polarization, for example). These differences are examined in light of surface scattering, cross calibration, and viewing geometry considerations.

Estimation of soil moisture. Analysis has focused on data collected during April, 1994 because of the favorable hydrologic conditions during SRL-1 (during the October, 1994 mission, the watershed was uniformly dry and no significant precipitation occurred). After registering the images, a land cover classification was created that was needed for both the radar inversion and the hydrological models. The land cover classification combined the algorithm of Pierce et al. (1994) (with classifications of urban, tall vegetation, short vegetation and bare soil) and the NDVI estimation from a Thematic Mapper image acquired April 1994. The NDVI allows us to distinguish active short crops (like winter wheat) from pasture and rangeland that are just coming out of senescence.

Following the land cover classification, two soil moisture estimation procedures using SIR-C data were applied: an empirical regression similar to that described in Lin et al. (1994) and the inversion model of Dubois et al. (1995). In addition, a distributed hydrological model was also used to estimate soil moisture within the Little Washita for the April SIR-C mission period. This allows for the intercomparison of the soil moisture estimates, and will help determine the usefulness of remotely sensed soil moisture for hydrological modeling. Ground-based soil moisture data from selected fields were used to help calibrate the hydrological model. The hydrological model was then used to simulate the entire watershed at a resolution of 30 m, using topography, soils, and meteorological data provided by USDA. In addition, the SIR-C data were used to help estimate the surface soil moisture conditions on April 11 prior to a major rain event. The period of simulation was from April 5 - 16, 1994. Comparisons between the model and ground data (field averaged soil moisture) are shown in Figure 1. Figure 2 shows a scatterplot of estimated field average soil moisture comparisons between ground based soil samples and the hydrological model (crosses) and between ground based soil samples and SIR-C-based estimates (open circles). These results show that the r² values and RMS errors of the two approaches are very similar--approximately 3% by volume. Finally, Figure 3 shows a time series of the watershed average soil moisture from the model with SIR-C soil moisture estimates at the SIR-C overpass times. Additional details are provided in Pauwels et al., 1997.

These results show that soil moisture estimates based on SIR-C inversion models and hydrological models give nearly equivalent results, and that SIR-C data could be useful in updating surface soil moisture in the hydrological model. The results also suggest that the hydrological model can accurately predict observed soil moisture for periods of 7 to 10 days without a loss of accuracy.

Scaling of soil moisture. The objective of this aspect of the project was to determine whether radar data can be used to estimate soil moisture at regional scales (10's to 100's km). At present, using the full resolution radar data would result in data volumes too large to allow estimation at these spatial scales. This research was initially carried out with remotely sensed soil moisture collected during the Washita '92 field experiment. Results published in Dubayah et al. (1995) indicate that the soil moisture exhibit multiscaling properties, namely log-log linearity of statistical moments as a function of scale and non-linear dependence of scaling exponents with order moment.

This work has been extended using the SIR-C radar data for April over the Little Washita and along a 150 km long ascending orbit strip for the April 12 acquisition. The important research issues regarding retrievals at this scale are: (i) if the radar data are 'scaled' or averaged up, then how does the retrieved large scale soil moisture field vary from that estimated by field from the full resolution radar data and then averaged; (ii) how can areas for which the radar soil moisture inversion algorithms are invalid (like high vegetation) be masked out as the radar data are averaged up to larger scales; and (iii) can large scale (averaged) radar data be used in a statistical manner to retrieve small scale properties, like small scale soil moisture variability, which are important to land-atmospheric exchanges of water and energy?

The project addressed these questions and successfully developed a procedure for estimating soil moisture at a range of scales, and for retrieving small scale soil moisture properties from large scale averaged radar signals. Figure 4 shows the approach used for the scaling analysis. The critical steps are determining how to mask out areas of high vegetation for which the radar inversion models are invalid. Figure 5 and 6 provide the scaling results for two 25 x 25 km blocks within the 150 km strip. In both figures, the x-axis represents the averaging scale starting at a 100 m pixel resolution and going up to 6.4 km resolution. Figure 5 shows that the mean soil moisture is scale invariant; i.e., if the vegetation is correctly masked, soil moisture estimated from radar data averaged up to 6.4 km grids has the same value as the mean soil moisture estimated from 100 m radar data (with all the appropriate vegetation masking) and then averaged up. The scaling work confirms earlier project research using the 1992 airborne data on the Little Washita as mentioned above. Figure 6 shows the scaling results for the variance (log-log linearity) of the dielectric constant (solid dots). The solid line shows the best fit of the log variance for scales larger than 800 m resolution, while the dashed line shows the estimated small scale variance extrapolated from the large scale analysis. The good agreement between the dashed line and solid dots demonstrates that the large scale (averaged) radar data can be used to retrieve small scale soil dielectric variance and therefore soil moisture variance. Further details are presented in Crow et al. (1996) and Crow et al. (1997). This is an important result, since it suggests that radar data can be used for regional mapping of soil moisture, and that the small scale variability can be estimated and used to determine the fractional areas of a large scale grid that are under either soil controlled or atmospheric-controlled evapotranspiration, which are critical parameters for macroscale water and energy balance modeling.

Incorporation of SIR-C derived soil moisture fields in hydrological modeling. The usefulness of incorporating SIR-C derived soil moisture information in the Princeton TOPLATS distributed hydrological model is being investigated for the Little Washita watershed in Oklahoma. Previous work using data from the Washita '92 experiment has demonstrated that initializing the hydrological model with surface soil moisture estimated from microwave remote sensing produced more accurate surface soil moisture fields compared to the more traditional initialization via streamflow/water table analysis. The accuracy of the model soil moisture predictions was maintained throughout an eight-day dry down after a single initialization on the wet day, illustrating that the model performance did not degrade with time over this period. After the remote sensing initialization, the model could also resolve small farm ponds and saturated areas in ways that were impossible with the streamflow initialization.

This analysis is currently being repeated using the SIR-C data from SRL-1 in April, 1994. In the first set of model runs, the hydrological model is initialized using standard streamflow/water table methodology on April 5 and allowed to run for the next thirteen days based on the meteorological forcings. In the second set of model runs, the model is started on April 5 as before but then is re-initialized on April 12 using surface soil moisture fields retrieved from the SIR-C radar data following the Shi modified IEM inversion approach (the first SIR-C data take over Chickasha on April 11 occurred during an active rain event and was therefore discarded for soil moisture intialization purposes in favor of the next data take on April 12). The model is then allowed to run through April 18 as before based on the meteorological inputs. In watershed areas not covered by the SIR-C data take on April 12, the surface soil moisture fields as calculated by the baseflow model are used (i.e., only those pixels with a SIR-C response and an IEM-derivable soil moisture are re-initialized).

Figure 7 shows surface soil moisture images for six days during the SRL-1 mission produced by the model after the SIR-C initialization on April 12. The drydown of the watershed after the rainfall event is very obvious. Figure 8 is a plot comparing the two different model initialization schemes with the ground measurements of surface soil moisture superimposed; all values are watershed averages of the 0-5 cm surface layer. During April 12-14 ground sampling was performed both in the morning and afternoon, and the diurnal soil moisture change is surprisingly large (especially on April 14). The model run using the standard baseflow initialization produces estimates of soil moisture which are consistently wetter than the model output with the SIR-C initialization, and only match the ground measurements during the mornings immediately after the rainfall. The model using the Shi SIR-C soil moisture from April 12 initially underestimates soil moisture, but then matches the afternoon ground measurements and subsequent drydown quite well.

Figure 8 also demonstrates the difficulty of modeling soil moisture variability throughout the day. Additional diurnal field data, in conjunction with basin-scale remote sensing, are needed to help improve the soil evaporation parameterization in hydrologic models. While strong diurnal variability has been observed previously at point-scale measurements, the strength of the diurnal signal observed at the Little Washita basin-scale, where there is a broad range of soil and vegetation types, is both surprising and a challenge for hydrologic modeling.

Another area of ongoing work is the improvement of radar soil moisture retrieval methods for vegetated areas, in particular the incorporation of vegetation effects in the IEM modeling approach, and documentation of the subsequent improvement in hydrological model performance.

REFERENCES, PUBLICATIONS AND PRESENTATIONS

Crow, Wade, Eric F. Wood and Ralph Dubayah, 1996, Regional Soil Moisture Estimation Through Scaling, Remote Sensing and Hydrological Modeling, Spring Meeting, American Geophysical Union, Baltimore, MD, May 1996.

Crow, Wade, Eric F. Wood and Ralph Dubayah, 1997, Spatial Scaling of SIR-C Radar Data and the Regional Estimation of Soil Moisture, in preparation.

Dubayah, R., E. F. Wood and D. Lavallee, 1995, The spatial scaling properties of remotely sensed and modeled near-surface soil moisture state for the Little Washita basin, in Quatrocchi D. and M. Goodchild, (editors), Scaling of Remote Sensing Data for GIS.

Dubois, P., J. van Zyl and E. Engman, 1995, "Measuring Soil Moisture with Imaging Radars", IEEE Trans. Geosc and Rem. Sens., 33, pp 915-926.

Fung, A., K. Chen, A. Hsu, E. Engman, P. O'Neill, and J. Wang, 1996, " A Modified IEM Model for Scattering from Soil Surfaces with Application to Soil Moisture Sensing," Proc. of IGARSS '96, IEEE, Lincoln, NE, May 27-31, 1996, Vol. II, pp. 1297-1299.

Jackson, T., L.L. Tang, A. Hsu, E. Wood, P. O'Neill, and E. Engman, 1996, Washita '94 SIR-C/

X-SAR Data Sets, published on CD-ROM by NASA Goddard Space Flight Center, Greenbelt, MD, March, 1996.

Jackson, T., L.L.Tang, E. Wood, A. Hsu, P. O'Neill, and E. Engman, 1996, "SIR-C/X-SAR as a Bridge to Soil Moisture Estimation Using Current and Future Operational Satellite Radars," Proc. IGARSS '96, IEEE, Lincoln, NE, May 27-31, 1996, Vol. II, pp. 1064-1066.

Lin, D.S., S. Saatchi and K. Beven, 1994, "Soil Moisture Estimation Over Grass-Covered Areas Using AIRSAR," International Journal of Remote Sensing, 15 (11), 2323-2343, 1994.

O'Neill, P.E., N.S. Chauhan, and T.J. Jackson, 1996, "Use of Active and Passive Microwave Remote Sensing for Soil Moisture Estimation Through Corn," International Journal of Remote Sensing, Vol. 17, No. 10, pp. 1851-1865.

O'Neill, P., A. Hsu, T. Jackson, and E. Wood, 1996, "Investigation of the Accuracy of Soil Moisture Inversion Using Microwave Data and Its Impact on Watershed Hydrological Modeling," Proceedings of IGARSS '96, IEEE, Lincoln, NE, May 27-31,1996, Vol. II, pp. 1061-1063.

O'Neill, P., J. Petrella, and J. Fuchs, 1996, "Chapter XVI: Multifrequency Truck-Mounted Radar System," in Hydrology Data Report: Washita '94, P. Starks and K. Humes, eds., NAWQL 96-1, USDA, June, 1996.

O'Neill, P. E., A. Y. Hsu, T. J. Jackson, E. F. Wood, and M. Zion, 1997, "Investigation of the Accuracy of Soil Moisture Inversion Using Microwave Data and Its Impact on Watershed Hydrological Modeling," Proceedings of the Third Intl. Workshop on Application of Remote Sensing in Hydrology, NHRI Symposium Series, Greenbelt, MD, October 16-18, 1996, pp. 211-226.

O'Neill, P., A. Hsu and T. Jackson, 1995, "Use of Active and Passive Microwave Measurements from Multiple Platforms for Soil Moisture Estimation", AGU Spring Meeting, Paper H32G-3, EOS, 76(17), April 25, 1995 (Supplement), p.5127.

O'Neill, P.E., A.Y. Hsu, and J.C. Shi, 1995, "Soil Moisture Estimation Using Time-Series Radar Measurements of Bare and Vegetated Fields in Washita '92," Proc. IGARSS '95, IEEE, Florence, Italy, July 10-14, 1995, Vol. I, pp. 498-500.

O'Neill, P., A. Hsu, and J.C. Shi, 1995, Soil moisture estimation using time-series radar measurements of bare and vegetated fields, Proc. IGARSS '95, IEEE, Florence, Italy, July 10-14, 1995, vol. I, pp. 498-500.

O'Neill, P.E., J.J. Petrella, and A.Y. Hsu, 1995, "Comparison of Multifrequency Truck Radar and SIR-C Backscatter for Soil Moisture Estimation in Washita '94," Proc. IGARSS '95, IEEE, Florence, Italy, July 10-14, 1995, Vol. I, pp. 368-370.

O'Neill, P.E., N.S. Chauhan, T.J. Jackson, D.M. LeVine, and R.H. Lang, 1994, "Microwave Soil Moisture Prediction Through Corn in Washita '92," Proceedings of IGARSS '94, IEEE, Pasadena, CA, August 8-12, 1994, Vol. III, pp. 1585-1587.

O'Neill, P. and J. Petrella, 1994, "Comparison of Truck-mounted Radar Measurements with SIR-C Microwave Data for Soil Moisture Estimation", NASA Research and Technology Report, GSFC, pp 124-126.

O'Neill, P.E., N.S. Chauhan, and R. Lang, 1993, "Ground-Based Radar Measurements During the Washita '92 Hydrology Experiment," AGU Spring Meeting, EOS Supplement, Baltimore, MD, May 24-28, 1993, p. S132.

Pauwels, V., E. Wood, and T. Jackson, 1997, A comparison of remote sensing and hydrological modeling to estimate soil moisture at the basin scale, Water Resources Research, under review.

Pierce, L. E., F. Ulaby, K. Sarabandi and M. Dobson, 1994, Knowledge-based classification of polarimetric SAR images, IEEE Trans. Geosc and Rem. Sens., 32(5):1081-1087.

Rogowski, Andrew S., and Edwin T. Engman, 1996, Using a SAR image and a decision support system to model spatial distribution of soil water in a GIS framework, Proc. Third International Conference on Integrating GIS and Environmental Modeling, Santa Fe, NM, January 1996.

Rogowski, Andrew S., Donald E. Simmons and Edwin T. Engman, 1997, Estimating soil moisture changes on an agricultural watershed, submitted to Water Resources Research.

Shi, J., J. Wang, A. Hsu, P. O'Neill, and E. Engman, 1995, "Estimation of Soil Moisture and Surface Roughness Parameters Using L-Band SAR Measurements," Proc. IGARSS '95, IEEE, Florence, Italy, July 10-14, 1995, Vol. I, pp. 507-509.

Shi, J.C., J. Wang, P. O'Neill, A. Hsu, and E. Engman, 1995, Application of IEM model to soil moisture and surface roughness estimation, Fifth Annual JPL Airborne Earth Science Workshop, JPL, Pasadena, CA, January 1995, vol. 3, pp. 51-54.

Shi, J., J. Wang, A. Hsu, P. O'Neill, and E. Engman, 1997, "Estimation of Bare Surface Soil Moisture and Surface Roughness Parameters Using L-Band SAR Image Data," IEEE Trans. on Geoscience and Remote Sensing, in press.

Wang, J.R., P.E. O'Neill, E.T. Engman, R. Pardipuram, J.C. Shi, and A.Y. Hsu, 1995, "Estimating Surface Soil Moisture from SIR-C Measurements Over the Little Washita Watershed," Proc. IGARSS '95, IEEE, Florence, Italy, July 10-14, 1995, Vol. III, pp. 1982-1984.

Wang, J.R., A. Hsu, J.C. Shi, P.E. O'Neill, and E.T. Engman, 1997, Estimating Surface Soil Moisture from SIR-C Measurements over the Little Washita River Watershed, Remote Sensing of Environment, vol. 59, pp 308-320.

Wood, E., P. O'Neill, T. Jackson, M. Zion, and A. Hsu, 1996, "Assessing Soil Moisture Remote Sensing Experiments for Hydrological Modeling and Analyses: Experiences from the Little Washita in 1992 and 1994," AGU Fall Meeting, Dec., 1996.

Wood, E. F., V. Pauwels, T. Jackson, E. Engman, P. O'Neill and F. Schiebe, 1995, "Using SIR-C Shuttle Radar Laboratory (SRL) for Soil Moisture Remote Sensing in the Little Washita Catchment: Initial Results", AGU Spring Meeting, Paper H32G-4, EOS , 76(17), April 25, 1995, pg S127.

Wood, Eric F., Ralph Dubayah and Dennis Lettenmaier 1996, Combining Hydrological Modeling and Remote Sensing for Large Scale Water and Energy Balance Studies, Spring Meeting, American Geophysical Union, Baltimore, MD, May 1996.

Zion, M., E. Wood, and P. O'Neill, 1996, "Use of Remotely Sensed Soil Moisture to Calibrate and Validate Hydrological Models, " Spring Meeting, AGU, Baltimore, MD, May, 1996.