WMOD00526 - Laserfiche WebLink

Home Browse Search

1338 JOURNAL OF APPLIED METEOROLOGY VOLUME 28 Bin positions 251 to 253 are used to indicate the bin positions of up to three project aircraft. A zero value in these positions indicates that no aircraft was detected. The presence of a number 1 to 250 indicates the range at which a project aircraft was located. For instance, if bin position 252 (frames 278,538, 798, and/or 1058) contains the number 37, this means the project aircraft designated as number 2 has been detected by a separate aircraft detection antenna to be located at bin position 37. Range-bin 37 would still contain a DVIP number which indicates an echo signal returned. The reflectivity value may be higher due to the existence of the aircraft at that location. 3. Computer processing software Most of the computer routines that make up the processing software are written in FORTRAN 77. Briefly, the processing procedure consists of 1) copying each raw record to disk, except those with bad record lengths (not 1062 bytes); 2) deleting nonecho (noise) data bins from this (raw-record) disk copy; 3) per- forming numerous quality and accuracy tests on the housekeeping information and a few tests on the data bins; 4) optimally compressing the raw data files; 5) comparing the compressed data files with the raw data files; and 6) producing a "first look" analysis product. Reports of these processing steps are printed. These include a listing of each error condition flagged (header edit report), a noise analysis report, and a comprehen- sive antenna-sweep summary report. The latter two reports along with the first-look analysis products (maximum reflectivity and echo top PPI displays) are also copied to microfiche. a. Raw data edit Original data can contain an assortment of errors. For example, date and/ or time may be set or incre- mented incorrectly; azimuth, elevation, and record count may be erroneous; or preset variables such as pulse repetition frequency and azimuth increment may have been reset by accident. When questionable con- ditions or errors are detected, they are flagged and noted in two locations. First, they are displayed in hard copy form in a header (housekeeping) edit report. Second, the flags are recorded into the permanent compressed archival file giving analysts the ability to filter, or isolate for special treatment, radials with any combination of flagged problems. A condensed summary of the tests performed on the data is listed in Table 1. The error or alert conditions tested are illustrated in seven broad categories to give the reader a general idea of the approach used. Most of the conditions tested in "Change Alert" are to alert the analyst to parameter changes that were expected to be stable. For instance, range delay would be ex- pected to stay constant through the day. Other flags not specifically shown in Table 1 also occur. For instance, if a parity error occurs while the raw tape is being read, a flag is appropriately inserted. A flag is inserted when a beginning of an azimuthal sweep is detected, or if a value is corrected. The only items that are allowed to be corrected in going from TABLE I. Summary of data quality and consistency tests performed. Digits 1-3 indicate the number of distinct conditions of a given type. I.D. code Year Day Hour Minute Second Azimuth Azimuth increment Elevation Elevation increment Range increment Range delay DVIP values Blue sky DVIP UN/LOG receiver Pulse repetition freq. Samples averaged Raw record count Range/height mode Required 0, S not all 0 Parity error count 3