Product Sheet ESCAT Soil Water Index 50 km Time by slappypappy129

VIEWS: 7 PAGES: 6

									Product Sheet: ESCAT Soil Water Index 50 km Time Series

The soil water index data (SWI) are derived from surface soil moisture data (SSM), which are
retrieved from the radar backscattering coefficients measured by the scatterometers onboard the
European Remote Sensing Satellites (ERS-1 and ERS-2) using a change detection method, developed
at the Institute of Photogrammetry and Remote Sensing (IPF), Vienna University of Technology (TU-
Wien). In the TU-Wien model, long-term scatterometer data are used to model the incidence angle
dependency of the radar backscattering signal σ 0 . Knowing the incidence angle dependency, the
backscattering coefficients are normalised to a reference incidence angle (40°). Finally, the relative soil
moisture data ranging between 0% and 100% are derived by scaling the normalised backscattering
coefficients σ 0 (40) between the lowest/highest σ 0 (40) values corresponding to the driest/wettest soil
conditions. Finally, a two-layer water model representing the topmost layer and the “reservoir” below
is applied, where the influence of SSM is modelled with an exponential function resulting in the soil
water index (SWI). This product represents the soil moisture content in the first 1 meter of the soil in
relative units ranging between wilting level and field capacity.

PARAMETER:             Soil water index (SWI)
SENSOR:                AMI-SCAT instrument (scatterometers onboard ERS-1/ERS-2 satellites)
PRODUCT
EXAMPLE:




                       Fig. 1: Soil water index (SWI) with error-bars for grid point index 852674 (location:
                       15.2577N, 8.611W)
RANGE AND              0-100 %
UNITS:
ACCURACY:              Each SWI value corresponds to a noise value, which shows the uncertainty
                       range
PERIOD:                1991-08-01 – 2007-05-31 (YYYY-MM-DD)
                       Note: data are not available in the period 2001-01-01 – 2003-08-12!
TEMPORAL               10-days. The temporal resolution is based on the specifications of the
RESOLUTION:            infiltration model used and is set to 10 days. The model integrates all
                       measurements from a 100-day interval prior to the SWI sample time and
                       weights the most recent measurements to contribute strongest to the
                       respective SWI sample. Data are stored for day 10/20/30 at 12:00 UTC for
                       each month
UPDATES:               Irregular, on demand
USAGE:                 Scientific datasets
DATA TYPE:             Time series in grid format

                                                        1
GRID TYPE AND Discrete Global Grid (DGG), which is an adapted sinusoidal grid using an
SIZE:         ellipsoid based on the GEM6 model. The grid is defined such that the
              spacing is approximately 12.5 km. In total, the grid consists of 3,264,391
              points, where 839,826 points are defined over land. Note that there are no
              data available over dense tropical forest, Greenland, Antarctica, sand desert
              areas, open water bodies, coastal areas and areas with high azimuthal
              anisotropy. The underlying satellite data are oversampled to the DGG grid.
              The grid is subdivided globally into 179 cells (Fig. 2). For more information
              about the grid definition see:

                    Kidd, R. (2005): Discrete Global Grid Systems. ASCAT Soil Moisture
                    Report Series, No. 4., Institute of Photogrammetry and Remote Sensing,
                    Vienna University of Technology.




                    Fig. 2: Cell structure of DGG.

                    The geolocation information of the grid points of the DGG is stored in the
                    two binary files "longitude.dat" and "latitude.dat", containing 3,264,391
                    elements.

                    Parameter:       longitudes of DGG points
                    File name:       longitude.dat
                    Data format:     binary
                    Data type:       float
                    No. elements:    3,264,391

                    Parameter:       latitudes of DGG points
                    File name:       latitude.dat
                    Data format:     binary
                    Data type:       float
                    No. elements:    3,264,391

                    Note:
                    Grid Point Indices (GPI) start with 0

                                                     2
              GPI=0       :the index of the first element in longitude/latitude array
              GPI=3264390 :the index of the last element in longitude/latitude array
SPATIAL       Coverage is global until 2001-01-01
COVERAGE:     Note: after this date, data coverage is limited only to the area of
              available ground receiving stations (Fig. 3)!




              Fig. 3: Geographic coverage of ERS receiving stations

              One example of the Discrete Global Grid (DGG) distribution over Austria is
              given in Fig. 4. The product is defined for each grid point, with a spacing of
              12.5 km.




              Fig. 4: Example of DGG distribution over Austria
SPATIAL       50 km, sampled to a grid spacing of 12.5 km
RESOLUTION:
KNOWN         1) The method to retrieve soil moisture is in principal a change detection
RETRIEVAL        method. Temporal variations can therefore be retrieved accurately,
PROBLEMS:        whereas the absolute level of soil moisture can be biased in certain
                 regions. Biased estimates are especially observed in extreme climates

                                             3
             (deserts, arctic regions).
          2) The azimuthal viewing geometry of the sensor is not taken into account
             during the retrieval. Azimuthal artefacts occur mainly in mountainous
             and sand desert regions.
          3) Retrieval of soil moisture is not possible under snow and frozen soil
             conditions or areas with highly undulated terrain.
          4) Open water surfaces are known to cause errors in the retrieval.

QUALITY   Following quality flags are deliverable with the data, giving an estimate for
FLAGS:    the problematic areas:
              1) Probability of snow covered land,
              2) Probability of frozen land,
              3) Inundation/wetland fraction,
              4) Topographic complexity.

          For more information see documentation:
          Scipal, K., V. Naeimi, S. Hasenauer (2005): Definition of Quality Flags.
          ASCAT Soil Moisture Report Series, No. 7., Institute of Photogrammetry
          and Remote Sensing, Vienna University of Technology.

          Quality flags are stored in binary files with following specifications:
          File names:
          XX_qflags.dat
          where:
          XX = cell number

          Each file includes binary values with the following structure:
          str = {        gpi    :long
                         snow :byte array(366)
                         frozen :byte array(366)
                         water :byte
                         topo :byte
                 }

          The snow/frozen arrays contain 366 values, representing the day of year
          (including February 29 for leap years).

          nan = 254 (Not-a-Number)

          1) Snow cover probability:
                0      Snow-free land
                1-100 Sea ice concentration (%)
                101 Permanent ice (Greenland, Antarctica)
                102-201 Snow (-101 => %)
                202-251 Not used
                252 Mixed pixels at coastlines (unable to reliably apply
                       microwave algorithms)
                253 Suspect ice value
                254 Corners (undefined)
                                       4
                      255 Ocean
               2) Frozen soil probability:
                      0-100 probability of frozen soil (%)
               3) Inundation/wetland fraction:
                      0-100 Water fraction (%)
               4) Topographic complexity:
                      255 Open water
                      0-100 Normalized standard deviation of heights in percentage
PROJECTION:    Coordinates are geographic (lon/lat)
DATA FORMAT:   Zipped raw binary (WinZip)
FILE NAMING    From 179 globally subdivisions of the DGG, 90 cells include land data. Time
CONVENTION:    series of SWI for each grid point are stored in separate files with the
               following format:
               TUW_ESCAT_SWI_W##_gpXXXXXXX.dat

               where:
               TUW=     processing facility (TU Wien)
               ESCAT=   sensor (ERS 1&2 scatterometer data)
               SWI=     data type (soil water index)
               W=       software (WARP processing software)
               ##=      version
               gp=      grid point index (GPI)
               XXXXXXX= index number

               Example:
               TUW_ESCAT_SWI_W50_gp2412903.dat
FILE SIZE:     Approximate data amount of global datasets:
               Soil water index:  1.6GB (in zipped format)
               Grid information:  28MB
               Quality flags:     30MB (in zipped format)
DATA           Each file includes binary encoded-values with the following structure:
ENCODING:
               Date: unsigned long number             (1st-4th byte)
               SWI: byte                              (5th byte)
               Noise: byte                            (6th byte)

               -   Date is encoded as a unsigned long number: YYYYMMDDHH (Y:year,
                   M:month, D:day, H:hour)
               -   To get the original values of SWI and Noise a multiplication factor of
                   0.5 has to be applied (e.g. value: 40, SWI: 40*0.5=20%)
               -   Decoded data range from 0 to 100. Values greater than 254 are NaN (not-
                   a-number).
EXAMPLE AND    Imagine you want to read soil moisture time series for a point located in
SAMPLE         Austria at 47.8°N, 13.5°E. You need to find first the nearest DGG point by
DATASET:       searching in the files “longitude.dat” and “latitude.dat”. In this example, the
               nearest neighbor is located at lon: 13.5246/lat: 47.8345, which is the
               2,412,903th element of longitude/latitude data arrays. It means that

                                           5
               GPI=2,412,903 is the index of the nearest grid point (considering gpi
               numbers start with 0). Knowing the cell number (97) and having the GPI,
               you can read the corresponding soil moisture data by reading the respective
               binary-encoded file name "TUW_ESCAT_SSM_W50_gp2412903.dat",
               located within the file “97.zip”. This will give you arrays of dates, SSM and
               noise.
               A sample dataset is available in the folder \sample_data. It contains
               procedures written in IDL-language to read the binary data and contains
               some sample ASCII outputs.
SOURCE:        IPF/TU WIEN
               Institute of Photogrammetry and Remote Sensing (IPF), Vienna University of Technology
               (TU WIEN), Austria. http://www.ipf.tuwien.ac.at/radar
DATA           SFTP (Secure File Transfer Protocol, with user name and password).
DELIVERY:      A smooth download is best done with secure ftp software to access the SFTP
               server, e.g. the open source software “WinSCP” (http://winscp.net/)
COST:          Free of charge for non-commercial use
PROCESSING     WARP 5.0, written in language IDL 6.3 (Windows platform)
SOFTWARE:
REFERENCES:    Selected references introducing retrieval method:
               1) Wagner, W., G. Lemoine, H. Rott (1999): A Method for Estimating Soil
               Moisture from ERS Scatterometer and Soil Data, Remote Sensing of
               Environment, Vol. 70, pp. 191-207. doi:10.1016/S0034-4257(99)00036-X
               2) Wagner, W. (1998): Soil Moisture Retrieval from ERS Scatterometer
               Data, PhD dissertation, Vienna University of Technology, Austria.

               Selected references for product validation:
               3) W. Wagner, K. Scipal, C. Pathe, D. Gerten, W. Lucht, B. Rudolf (2003)
               Evaluation of the agreement between the first global remotely sensed soil
               moisture data with model and precipitation data, Journal of Geophysical
               Research - Atmospheres, Vol. 108, No. D19, 4611,
               doi:10.1029/2003JD003663
               4) Ceballos, A., K. Scipal, W. Wagner, J. Martinez-Fernandez (2005)
               Validation and downscaling of ERS Scatterometer derived soil moisture
               data over the central part of the Duero Basin, Spain, Hydrological Processes,
               vol. 19, pp. 1549-1566, doi:10.1002/hyp.5585

               See full publication-list:
               http://www.ipf.tuwien.ac.at/radar/index.php?go=a_publications
DATA ACCESS    Please fill in the form: http://www.ipf.tuwien.ac.at/radar/index.php?go=data
AND CONTACT:
               For assistance with the datasets, please contact microwave@ipf.tuwien.ac.at
DOCUMENT       2008-10-13
UPDATE:
REVISION:      Version 2.3
DOCUMENT       \\global\swi_ts_w5\TUW_ESCAT_SWI_W50_product_sheet.pdf
URL:

                                              6

								
To top