The principal task of rainfall preprocessing is to transfer daily rainfall data from the original monitoring table DAILY_RAINFALL (source table) into the destination table Jrain which is more suitable for modelling purposes.
| AGENCY
_CODE |
STATION
_ID |
READING
_DATE |
DAILY_
RAINFALL |
CLIMATE_
REMARKS_ CODE |
COMMENTS | DATE_
LAST_ UPDATED |
USER_NAME |
|---|---|---|---|---|---|---|---|
| WAJ | AD0032 | 01-Jan-94 | 40.4 | - | 29-Apr-98 | JORDAN | |
| WAJ | AD0032 | 02-Jan-94 | 15.7 | - | 29-Apr-98 | JORDAN | |
| WAJ | AD0032 | 03-Jan-94 | 40 | - | 29-Apr-98 | JORDAN | |
| WAJ | AD0032 | 04-Jan-94 | 33.8 | - | 29-Apr-98 | JORDAN | |
| ... |
Table DAILY_RAINFALL does not contain any records with zero rainfall (except for the days specially marked as data gap or accumulated reading). Each record has a "CLIMATE_REMARKS_CODE" with the following meanings:
| Code | Meaning |
|---|---|
| - | ordinary reading |
| e | estimated value |
| a | day with accumulated reading |
| c | start of data gap |
| d | end of data gap |
The processing from the source to the destination table has the purpose to obtain continuous data sets, i.e. also days with zero-rainfall are getting a record. Furthermore, cumulative periods (i.e. a single reading after a sequence of days with rainfall) are filled with daily data and data gaps are cut out.
In further subroutines, annual totals from totalizer stations are converted into daily values and socalled "virtual stations" are introduced in order to improve spatial data density in border areas.
Note: graphic is "clickable" and leads to related page