Documentation for Draper and Long's QuikSCAT/SeaWinds HDF Rain product files David G. Long Brigham Young University long@ee.byu.edu (801) 422-4383 last revised: 01/01/2004 DGL BYU Level 2 rain (L2R) files are "overlay" files for standard QuikSCAT/SeaWinds L2B files. They are generated from the standard JPL-produced L2A and L2B files and have an organization which mimics the L2B HDF file organization. L2R files contain winds estimated using Draper and Long's simultaneous wind/rain retrieval algorithm described in D.W. Draper and D.G. Long, "Simultaneous Wind and Rain Retrieval Using SeaWinds Data," in review, IEEE Trans. Geosci. Rem. Sens., 2003 and D.W. Draper and D.G. Long, "Assessing the Quality of SeaWinds Rain Measurements," in review, IEEE Trans. Geosci. Rem. Sens., 2003. The rain/wind model is described in D.W. Draper and D.G. Long, "Evaluating the effect of rain on SeaWinds scatterometer measurements," to appear, JGR, 2004. A typical L2R file contains the following HDF structure: HDF file format: (from hdfinfo) Number of datasets = 15 Number of attributes = 9 Dataset/Name Rank Dimensions Scale Offset Type ------------------------------------------------------------------------------------ 0 wvc_row 1 1624 0 0 1.00000 0.00000 int*2 1 wind_speed 3 4 76 1624 0.01000 0.00000 int*2 2 wind_dir 3 4 76 1624 0.01000 0.00000 uint2 3 rain_rate 3 4 76 1624 0.01000 0.00000 int*2 4 max_likelihood_est 3 4 76 1624 0.00100 0.00000 int*2 5 num_ambigs 2 76 1624 0 1.00000 0.00000 byte8 6 wvc_selection 2 76 1624 0 1.00000 0.00000 byte8 7 percent_rain 3 4 76 1624 0.01000 0.00000 int*2 8 wind_speed1 3 4 76 1624 0.01000 0.00000 int*2 9 wind_dir1 3 4 76 1624 0.01000 0.00000 uint2 10 num_ambigs1 2 76 1624 0 1.00000 0.00000 byte8 11 wvc_selection1 2 76 1624 0 1.00000 0.00000 byte8 12 regime 3 4 76 1624 1.00000 0.00000 byte8 13 wvc_selection_opt 2 76 1624 0 1.00000 0.00000 byte8 14 set_selection_opt 2 76 1624 0 1.00000 0.00000 byte8 15 wvc_quality_flag 2 76 1624 0 1.00000 0.00000 int*2 16 rain_confidence_flag 2 76 6 0 1.00000 0.00000 byte8 ------------------------------------------------------------------------------------ global attributes: Index Attribute Name Values -------------------------------------------------------------------------------- 0 LongName QuikSCAT Level 2B Ocean Wind Vectors and Rain in 25km Swath 1 ShortName QSCATL2R 2 producer_institution Brigham Young University Microwave Earth Remote Sensing (MERS) Laboratory 3 InstrumentShortName SeaWinds 4 PlatformLongName NASA Quick Scatterometer 5 PlatformShortName QuikSCAT 6 data_format_type NCSA HDF 7 L2Rfilename QS_S2R03221.20001592043 8 L2Afilename QS_S2A03221.20001592043 9 L2Bfilename QS_S2B03221.20001592046 10 WindModel QS_MODL0003 11 RainModel BYU V6 Quadratic 12 RainThresholds BYU_rainflag_thresholds_V0.txt 13 Investigator David G. Long, long@ee.byu.edu (801) 422-4383 14 build_id 1.000/2004_01_01 -------------------------------------------------------------------------------- The contents are desdribed in the following The wind_speed, wind_dir, max_likelihood_est, and num_ambigs array are similar to the arrays of the same name from the corresponding L2B file, but contain the wind/rain ambiguities. Simultaneous wind retrieval is not done for the outer (V-pol only) swath. In this case the rain_rate is set to zero and the reported wind ambiguities are from wind-only retrieval. (Thus, in the far swath the wind_speed and wind_speed1 values are identical.) Also, the wind-only retrieval is reported if the average wind speed of the wind-only ambiguities is less than 4 m/s. Note that wvcs with high rain rates may result in zero wind speed ambiguities. The wind_speed1, wind_dir1, and num_ambigs1 are the wind-only retrieval wind ambiguities and are copies of the L2B wind ambiguities. The various data sets contain: wvc_row: (copy of L2B wvc_row variable) wvc row number wind_speed: wind speed ambiguities (m/s), in MLE value order wind_dir: wind direction ambiguities (deg), in MLE value order rain_rate: rain rate ambiguities (km mm/hr), in MLE value order max_likelihood_est: MLE value from wind or wind/rain retrieval num_ambigs: number of ambiguities at each wvc (max=4) wvc_selection: index (1..4) of selected sim wind/raind ambiguity based on median-filter-based ambiguity removal percent_rain: inferred wvc rain fraction (percent) for each ambiguity wind_speed1: wind-only wind speed ambiguities (m/s), in corresponding MLE value order wind_dir1: wind-only wind direction ambiguities (deg), in corresponding MLE value order num_ambigs1: number of wind-only ambiguities at each wvc (max=4) wvc_selection1: index (1..4) of selected wind-only ambiguity based on median-filter-based ambiguity removal (copy of L2B wvc_selection) regime: estimated wind/rain retrieval regime 0 - Estimated rain does not significantly affect wind (rain estimate poor) 1 - Estimated rain and wind backscatter are of same order (both rain and wind estimate OK) 2 - Estimated rain dominates wind (wind estimate quality poor) wvc_selection_opt: ambiguity selection index (1..4) from combined simultanedou wind/rain and wind-only retreival. Index is for simultaneous wind/rain ambiguities (wind_speed) where set_selection_opt=0 or from wind-only ambiguities (wind_speed1) where set_selection_opt=1 set_selection_opt: flag to indicate which retrieval method is used for each wvc =0 for simultaneous wind/rain retrieval, =1 for wind-only retrieval wvc_quality_flag: copy of L2B wvc_quality_flag rain_confidence_flag: rain estimate quality confidence flag 0 - Low confidence in estimated rain 1 - High confidene in estimated rain