SITECHAR SITEMAPS GPHOTOS SOILMOIS NS001 TM ERS-1 SIRCAPR SIRCOCT JPLSCALE TKRADAR Washita'94 SIR-C/X-SAR Data Sets

T. Jackson, USDA ARS Hydrology Lab;
L. Tang, Chinese Remote Sensing Satellite Ground Recieving Station, Beijing, China;
A. Hsu, SSAI, Lanham, MD 20706;
E. Wood, Dept. of Civil Engineering, Princeton Univ. Princeton, NJ;
P. O'Neill and E. Engman, NASA Goddard Space Flight Center, Laboratory for Hydrospheric Processes, Hydrological Sciences Branch, Greenbelt, MD 20771;
ver. 10/18/95

This CD contains a variety of data sets collected as part of the Washita 94 experiments. The focus of this effort is on the SIR-C/X-SAR mission. A data report is available that describes the other investigations that were conducted as part of Washita 94.

There were two SIR-C/X-SAR missions: April and October. During each mission data were collected in one form or another on almost every day, sometimes ascending and descending passes on the same day.

Contents

Directory SITECHAR GPHOTOS SOILMOIS NS001 TM ERS-1 SIRCAPR SIRCOCT JPLSCALE TKRADAR

Description Descriptions of the ground test sites Site location maps Ground photos of selected sites Average soil moisture for each site Aircraft false color composite images collected 4/6/94 Landsat Thematic Mapper data from 4/12/94 SIR-C and X-SAR data from April (dB) SIR-C data from October (dB) SIR-C contrast stretched images L and C band backscatter from truck radar

SITECHAR

This directory contains data files of the site characteristics by month. The majority of the test sites used in Washita'94 were selected to provide representative spatially distributed soil moisture information. Several sites were used solely for transect soil moisture studies and a few other sites were sampled occasionally by investigators for specific reasons. In addition, some changes in the sites used were made during the three different experimental periods. The purpose of this section is to summarize the site characteristics. Summaries of the conditions are presented in files SCAPR.WK1 and SCOCT.WK1 (also as ASCII text files *.TXT). The methods and codings are described in the following sections.

A. Site Coding

Each site is described by a two number code. The first number represents the area and the second the site number within the area.

B. Soil Texture

The soil texture data presented in file SOILTEX.WK1 are for the surface 5 cm layer. These values are based on laboratory particle size analysis conducted at the USDA Hydrology Lab as part of the current study.

C. Bulk Density

The method used in this investigation is a volumetric displacement procedure that has been successfully employed in previous experiments. The procedure uses a specially designed bulk density ring with a hook gage and securing bolts. The ring is placed on the ground and secured by driving the bolts into the soil. A plastic film is then placed inside the ring. Water from a graduated cylinder is then used to measure the background volume. After removing the water and plastic, soil is extracted to a specific depth, in this case 5 cm. This soil was placed in plastic cooking bags and sealed. The plastic is then returned to the ring and the total volume is measured using water from the graduated cylinder. The volume of soil extracted is the difference between this volume and the initial background volume (values typically run between 700 and 900 ml). The soil sample is then returned to the lab where a wet weight is obtained. It is then oven dried and weighed again for a dry weight. The bulk density of the soil is computed by dividing the dry weight (less any tare) by the soil volume. The result is in g/cm3.

We attempted to obtain four samples from each soil moisture grid sampling site. In some cases the are no samples, however, it is possible using field observations to use representative values for fields that were not sampled. The field averages are listed in the data files. The values appear to be consistent with those we would expect from previous experience. Additional details on individual samples and conditions are available in the field notebooks.

D. Vegetation Parameters

Vegetation wet and dry biomass were determined by sampling two 18 x 18 inch areas in each field. For the first experimental period, most of the vegetation samples were obtained between April 7 and April 9 during dry soil condition. After the rain of April 11, some growth in the winter wheat was observed. Two fields (13 and 31) were sampled again on April 15.

E. Surface Roughness

Several photographs of a gridded board inserted into the soil surface were obtained for each sampling site. In some cases, it was quite difficult to place the board correctly due to the soil's hardness. In other situations, the vegetation was deemed to be too dense for photographing the soil surface.

These photos were enlarged and a digitizer was used to measure the heights of the surface (approximately 100 points per photo). These x and y values were then used to compute the standard deviation of the surface heights and the correlation lengths listed in the files.

SOILMOIS

This directory contains data files of the field average gravimetric soil moisture by site for each day in April and October. The following information is from the Washita 94 Data Report in preparation.

A. Objectives

The primary purpose for collecting ground observations of surface soil moisture was to provide calibration and verification data for the microwave remote sensing instruments. Sites were selected in an attempt to include a representative sampling of the various soil textures and land cover conditions. In addition, it was desirable to have these sites spatially distributed over the study area (in case of spatially distributed rainfall events) and also to have some clustered for radar analysis.

B. Sampling Schemes

For most of the sites, the primary goal was to determine the field average soil moisture. Data were not geolocated to an accuracy suitable for point by point comparisons. It was intended that the average of all samples from a site be compared to the average of the microwave data. Within each field, data were collected on a grid basis with a sample every 100 m (typically 14 samples per site). Some of the sites had unusual geometry that required some modifications. In all cases, the original data notebooks contain diagrams for each site and date.

Other sites were sampled for transect analyses. In these fields, the samples were collected along a line or transect perpendicular to the main drainage. These values should be averaged with caution because there were definite biases in their locations.

C. Sampling Procedure

A gravimetric soil moisture sample was collected using a specially designed scoop that extracts a 5 cm cube. This sample is placed in a metal can and sealed. The can number is then recorded on a schematic map of the site showing its approximate location. In addition, several fields were also sampled for the 1 cm soil moisture.

These gravimetric samples were weighed at the ARS lab to obtain a wet weight. All samples were then dried for 24 hours and weighed again to obtain a dry weight. The gravimetric soil moisture was computed by dividing the difference between the wet and dry weight by the dry weight less the can tare.

D. Results

These gravimetric samples were weighed at the ARS lab to obtain a wet weight. All samples Files GSMAPR.WK1 and GSMOCT.WK1 summarize the site average and standard deviation of soil moisture for April and October (also as ASCII text files *.TXT). These values are volumetric soil moisture computed by multiplying the gravimetric soil moisture by the soil bulk density

E. Soil Temperature Sampling

These gravimetric samples were weighed at the ARS lab to obtain a wet weight. All samples Soil temperatures were collected at depths of 5 cm and 15 cm using metal dial type thermometers as well as digital probes. This was done only when there was a C-130 flight (April only) and only in selected fields. Soil temperature was collected at each gravimetric sampling location. All of the observations were recorded in field notebooks and field averages are listed in file STAPR.WK1 (also as ASCII text files *.TXT).

SITEMAPS

These gravimetric samples were weighed at the ARS lab to obtain a wet weight. All samples This directory contains TIF images (gray scale) of the general study region indicating the location of the 5 areas used for ground sampling. For each area there is a TIF file of the boundaries of the sites. These are based on 1:24,000 USGS topographic maps. (TIF files were generated from Aldus Photostyler)

NS001

Just prior to the April SIR-C mission, the NASA C-130 was flown over the region, acquiring data with a number of different sensors including the NS001. The NS001 multispectral scanner operates in the seven Landsat-D Thematic Mapper bands plus a band from 1.13 to 1.35 mm. The nominal bandwidths are as follows:

Band 1 2 3 4 5 6 7 8

Spectral bandwidth (mm) 0.458 - 0.519 0.529 - 0.603 0.633 - 0.697 0.767 - 0.910 1.13 - 1.35 1.57 - 1.71 2.10 - 2.38 10.9 - 12.3

Sensor specifications are:

IFOV: Ground Resolution: Total Scan Angle: Swath Width: Pixels/Scan Line:

2.5 mrad 7.6 meters at 3000 meters 100 deg 7.2 km at 3000 meters 699

For each of the study areas, coverage from April 6 was extracted from the NS001 data set. These data wete at various scales and only minimal geometric modifications have been made. Data from bands 2,3 and 4 were composited as a false color blue-green-red image for each area. These data have been stretched in the same manner that Landsat data are often presented and should be useful in interpreting land cover and vegetation. Quantitative analysis of these data would require more extensive atmospheric and geometric corrections (see the Washita'92 data report for more information). Locations for the truck radar observations are indicated on these images.(TIF files were generated from Aldus Photostyler)

TM

The Landsat TM data used here were acquired by USDA ARS Durant, OK. The full image was georegistered by USDA ARS Hydrology Lab to USGS topographic maps and then the primary study area was extracted. The data are in individual TIF grayscale files for each channel. (TIF files were generated from Aldus Photostyler)

The image was acquired April 12, 1994. The processed files contain 934 lines by 1467 pixels at a resolution of 30 m. Geolocation information is as follows:

Output Georeferenced Units UTM 14 S E000 Projection Universal Transverse Mercator Zone 14 S Earth Ellipsoid Clarke 1866 (NAD 27) Upper Left Corner 562000 E 3875000 N Upper Right Corner 606010 E 3875000 N Image Centre 584005 E 3860990 N Lower Left Corner 562000 E 3846980 N Lower Right Corner 606010 E 3846980 N

Lon Lat Upper Left Corner 98d19'13.57" W 35d01'03.09" N Upper Right Corner 98d50'17.25" W 35d00'49.92" N Image Centre 98d04'50.45" W 34d53'22.61" N Lower Left Corner 98d19'21.05" W 34d45'53.51" N Lower Right Corner 97d50'30.04" W 34d45'40.46" N

File names have the following form: TM412#.tif where # is the band number Band Wavelength Range (mm) 1 0.45-0.52 2 0.52-0.60 3 0.63-0.69 4 0.76-0.90 5 1.55-1.75 6 10.4-12.5 7 2.08-2.35

ERS-1

A single ERS-1 image was available for April 15, 1994 and was acquired by Princeton University. ERS-1 is a C band VV polarization radar operating at a nominal angle of 23 degrees. The image was acquired at the Canadian receiving station and processed as an SGF product. Due to a problem with the antenna correction applied to these data in processing, the USDA ARS Hydrology Lab had to correct the image using a program supplied by the Canadians.

The ERS-1 image was georegistered to the TM image using control points. The same TM subset was then extracted resulting in a TIF gray scale file of 934 lines by 1467 pixels with a resolution of 30 m. To extract the backscattering coefficient use the following formula:

dB = (DN/10.2) - 25

SIRCAPR

The original SIR-C data sets as processed and calibrated by NASA JPL were assembled by NASA GSFC Hydrological Sciences Branch. A description of the data sets available from April 1994 is listed below.

Date Ascending or Descending Incidence Angle at Center (deg.) Polarizations Processing X-SAR 4/11/94 a 28.0 hh hv vv slc y 4/12/94 a 42.3 hh hv vv slc y 4/13/94 a 50.1 hh hv vv slc y 4/14/94 a 56.3 hh hv slc y 4/14/94 d 48.3 hh hv slc y 4/15/94 a 60.2 hh hv vv mlc n 4/15/94 d 42.4 hh hv vv slc y 4/16/94 d 36.2 hh hv vv slc y 4/17/94 d 30.9 hh hv vv slc n 4/18/94 d 26.5 hh hv vv slc y

NASA JPL software was modified to output the data as a scaled backscattering coefficient. To extract this from the TIF gray scale files use the following formulas

C band hh and vv dB = (DN/(255/25)) - 25 C band hv dB = (DN/(255/25)) - 35 L band hh and vv dB = (DN/(255/35)) - 35 L band hv dB = (DN/(255/35)) - 45

These data sets were then georeferenced for each day to the TM image using control points. TIF gray scale files were named as follows

S4ddtbpp.TIF where

• dd = the day (11 through 18)
• t = ascending (a) or descending (d)
• b = C or L band
• pp = polarization (hh, hv or vv)

The primary study area was extracted from the original data and consists of 934 lines by 1467 pixels with a resolution of 30m. In addition, the X-SAR X band vv data were georegistered and processed as part of the above procedure (b=X); the extraction equation for X band vv is

dB = (DN/(255/25)) - 25

SIRCOCT

These data were processed in the same manner described for April. The data sets included here are listed below. No X-SAR data was available at the time of the preparation of these data sets.

Date Ascending or Descending Incidence Angle at Center (deg.) Polarizations Processing 10/02/94 a 28.4 hh hv vv slc 10/03/94 a 41.3 hh hv vv slc 10/04/94 a 49.9 hh hv vv slc 10/05/94 d 47.2 hh hv slc 10/06/94 d 40.1 hh hv vv slc 10/10/94 d 34.4 hh hv slc

These data sets were then georeferenced for each day to the TM image using control points. TIF gray scale files were named as follows

S0ddtbpp.TIF where

• dd = the day (11 through 18)
• t = ascending (a) or descending (d)
• b = C or L band
• pp = polarization (hh, hv or vv)

The primary study area was extracted from the original data and consists of 934 lines by 1467 pixels with a resolution of 30 m. NASA JPL software was modified to output the data as a scaled backscattering coefficient. To extract this from the TIF gray scale files use the same formulas described for April except C band hv on 10/5/94 and 10/10/94 dB = (DN/(255/25)) - 30

JPLSCALE

As mentioned in the description of SIRCAPR, the JPL software produces a scaled output designed to maximize the contrast in each band and polarization on each day individually. The output is a scaled DN of amplitude and the program outputs the scale factor. To compute backscattering from this data the following operation must be performed

Amplitude = (DN)/Scale Factor Power =(Amplitude)**2 dB = 10log10Power

For every file in SIRCAPR and SIRCOCT the scaled images were also created and georegistered. The filenames are structured the same except an O replaces the S .

TKRADAR

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 Washita '94 experiment during the April, 1994 Shuttle mission (funding limitations prevented its return for the August and October campaigns). Designed and operated in conjunction with George Washington University, the truck radar consists of L, C, and X band radars at frequencies of 1.6, 4.75, and 10 GHz. External field calibration for both like and cross polarization channels was accomplished by recording the truck radar response to a target of known backscatter, in this case a large aluminum dihedral corner reflector (side length of 76 cm) mounted on a rotating wooden base. Cross-polarization data sensitivities of 30 dB above the noise floor at L band and 15 dB above the noise floor at C band were achieved. The X band data were marginal at incidence angles of 40 and 50 degrees due to noise floor problems.

On each of eight days of data collection in April the truck radar took measurements at three frequencies (L, C, X), four polarizations (HH, HV, VH, VV), and three angles (30, 40, 50 deg) for four test fields which each represented a typical surface cover in the watershed (alfalfa #11, new corn/bare #12, winter wheat #13, and pasture #62). Boom height was 40 ft (12.2 m). Average radar backscatter values were obtained by sweeping the radar boom 120 degrees in azimuth across the target fields. 46 data points were collected at each polarization during each 4-minute azimuthal sweep. A new sweep was required for each desired incidence angle and frequency; all sweeps were repeated at L band in order to give 92 samples for averaging.

Calibrated backscatter data for each test field are in a separate file sorted by frequency by angle by date. The first two digits of each file name represent the identification number assigned to each field. Only L and C band data are included here due to the marginal nature of some of the X band data. Additional information about the truck radar system may be found in Chapter XVI of the Washita'94 data report. Observation locations are indicated on the NS001 images.

Points of Contact

For information about the Washita'94 SIR-C/X-SAR Data Sets contact

Thomas J. Jackson
USDA ARS Hydrology Lab
Bldg. 007, Rm. 104, BARC-West
Beltsville, MD 20705

tjackson@hydrolab.arsusda.gov
301-504-8511 (voice)

For information about or assistance in using SGP97 data at the DAAC, contact

Hydrology Data Support Team
EOS Distributed Active Archive Center (DAAC)
Code 610.2
NASA Goddard Space Flight Center
Greenbelt, Maryland 20771

hydrology-disc@listserv.gsfc.nasa.gov
301-614-5165 (voice)
301-614-5268 (fax)

---

Last update:Tue Feb 2 09:06:05 EST 1999 Page Author: Hydrology Data Support Team -- hydrology-disc@listserv.gsfc.nasa.gov Web Curator: -- Website Curator: Anthony Drake NASA official: Steve Kempler, DAAC Manager -- Steven.J.Kempler@nasa.gov