Document title: Description of processing done at NSSDC for the NDADS DE datatype WATS_2S_ASCII. Project: DE NDADS Datatype: WATS_2S_ASCII EID: DOCUMENT Super-EID: DOCUMENTS There may be other documents also identified by this super-EID. NDADS filename: wats_nssdc_pro_de.doc TRF entry: b47608.txt in NSSDC's controlled digital document library, Dec. 1998. Document text follows: ---------------------- DE-2 Dynamics Explorer 2 WATS Wind and Temperature Spectrometer The Original WATS High-Resolution Data and Software: ---------------------------------------------------- The original PI-provided WATS high-resolution data were in VAX/VMS binary format with one file per orbit. This data are archived at NSSDC on magnetic tape (81-070B-04C). In addition the PI team had provided a FORTRAN program WATSCOR to correct the WATS data. WATSCOR makes the following corrections that are recommended by the PI team: (1) Temperature measurements following a RPA-mode measurements are not reliable and are therefore set equal to the fillvalue 9999. (2) Increases the temperature measured during horizontal baffle postion by 1% and during vertical position by 5%. (3) Decreases oxygen temperature by 3%. (4) Averages out the sweep assymetry for the horizontal or vertical neutral winds in geographic coordinates. Processing Done at NSSDC: ------------------------- I. As recommended by the PI, we first corrected the data with the WATSCOR program, and include the corrected and uncorrected temperatures and velocities in the output file. To run the WATSCOR program the following changes were necessary and were implemented before creating the WATSCOR-corrected orbit files: (1) Some of the array sizes had to be increased so that the program could handle the files from the whole DE-2 mission. (2) It was found that what is described as varibale 9 in FORMAT.SFD is really variable 10 and vice versa. The decoding of the packed variables 9 and 10 was done with the FORTRAN routine IBITS. (3) In subroutine VELCOR the last two 'IF ...'AND' statements contain the uninitialized variable IE, which should be really II. II. The following additional data processing was done at NSSDC (program OAWATSREAD.FOR): (1) The corrected and uncorrected temperatures (new words 8 and 9) and velocities (new words 15 and 16) were included in the ASCII output file. (2) Several of the instrument mode and status parameters were stored as bits packed into the original VAX/VMS binary words. To make access to these parameters easier we have decoded all of these parameters and included them as separate words in the output file: parameter packed in old word new word SLOT 10 (NOT 9 as indicated 4 in the SFDU docs) TIME1 10 13 TIME2 10 14 COUNTS1 9 (NOT 10 as indicated 11 in the SFDU docs) COUNTS2 9 12 (3) Variable values outside the valid range for the specific variable were replaced by the following fill values: word(s) description valid range fill value ------------------------------------------------------------------ 4 slot position 1,2,3,4 9 5 baffle direction 0,1 9 6 mass in AMU 0 < mass < 100 -9 8,9 temperature [K] 200 <= Tn <= 4000 -9 10,15,16 velocity [m/s] 4000 <= v <= 4000 9999 11,12 counts 0 <= counts < 100000 -9 13,14 baffle times 0 <= counts < 100000 -9 23(o/a) L-value 0 < L < 100 -9 There were also two records that had an invalid date (a date of 1183894048 during 82255 and a date of 0 during 82351) and were discarded. (5) The following orbit parameters were added using the OAREAD Fortran subroutine and the OABASE binary o/a data base: word description units ----------------------------------------- 17 Orbit number 18 Altitude [km] 19 Geographic latitude [degree] 20 Geographic longitude [degree] 21 Local Solar Time [hours] 22 Local Magnetic Time [hours] 23 McIllwain L value 24 Invariant latitude [degree] 25 Solar Zenith Angle [degree] (3) The orbit files (one per orbit) were combined to daily files for easy access to all the data for a specific day of year.