Revision as of 09:56, 1 September 2006; view current revision
←Older revision | Newer revision→
| Partly by R. Gommes |
Understanding Water Balance Output
Examining the water balance output information generated.
When the water balance calculations have been finalized 5 output screens are presented. All 5 are explained below.
Summary information per station.
Some story
Water Balance details per station.
Some story
Rangeland index per station.
Some story
Graph for one station.
Some story
Index distribution for all stations.
Some story
Examining the detailed output files.
All output is written to 7 files, of which the first is in a plain text format. It can be read with any editor or word-processor and be easily printed.
(1) The text output file, ccWsyyOX.TXT (or cc2syyOX.TXT when the current year is post 1999) contains details of the water balance calculations, station by station and dekad by dekad. A sample TXT output is given as annex 7.
The other output files are in DAT format. They all contain the following variables, next to the station name:
Id, the crop identification number (see 3.1 above: maize=1,... ETP=8). ETP is a "dummy" crop with no phenology (the crop coefficient stays permanently at 1.0);
X, Y, Z are longitude, latitude, and elevation. As these parameters are given for reference only (they enter no calculations) they can be given as "missing" in the input files without any problem to FAOINDEX. But they will probably be needed in the further processing of the files and, therefore, it is good practice to use correct values.
The additional variables which occur in each of the output files are described hereunder together with the description of files (2) to (7).
(2) the summary output file ccWsyyXA.* (or cc2syyXA.*) lists the parameters which are likely to be used for further processing and analysis, in particular for crop forecasting. The remaining files are more "monitoring-oriented" as they contain sequences of several dekad values for display and inter-station comparisons (for instance with ADDATI). The variables listed in the summary are the following:
H and WHCi, the soil Water Holding Capacity, and its initial value, respectively;
E and EfR%, the ratio between effective and actual total and rain, in percent;
P, the planting dekad (1-36);
C, the length of the crop cycle (dekads);
TWR, total water requirements in mm;
Indx, IndxNor and IndxLast, FAO soil water satisfaction index, in % of requirements, respectively actual, average and "last" values. Actual corresponds to the value expected at the end of the cycle, average is the end-of-cycle index computed with average rain and PET, all other parameters being those from the crop file; the "last" value is that of the last dekad for which actual data are available ();
EXWT and DEFWT: excess and deficit water totalled over the growing cycle;
ETA: actual crop evapotranspiration (i.e. actual crop water consumption), a factor directly related with crop yield;
%Av, ("available") the percentage of actual rainfall data used for the calculations (the rest being assumed to be average)
Cr1 to Cr4 and Ncr1 to NCR4 (for "Crossing" and "Normal Crossing") indicate the dates (dekads) when the rangeland index (see (7) below) crosses the 0.4*PET line. The dekads given in the file can be used to map the beginning of the actual (current) season (Cr1 to Cr4) and the normal beginning (NCr1 to NCr4). In areas with only one rainfall peak, there is only one normal start, but there are two in bimodal rainfall areas. In marginal areas, where rainfall is very close to 40% of PET, there may be more than one start even in areas with unimodal rainfall. The actual starts may be more than the normal starts, particularly in the not so rare event of false starts when rain decreases again after a first start: in the field, farmers must frequently replant when this happens.
(3) ccWsyySA.* (or cc2syySA.*) contains the soil water data, computed for each dekad, in the rows of the file. Note that dekads are crop dekads (from 1 to cycle length). Values beyond cycle length are coded with the default value for missing data. D1,D2...D36 are the dekad numbers in crop dekads (i.e. planting dekad is 1 and the last values corresponds to C, the cycle length).
(4) ccWsyyTA.* has the same structure as file (3); it contains the estimated crop actual evapotranspiration for each dekad.
(5) ccWsyyIA.* list the values of the FAO water satisfaction index.
(6) Water surplus in mm (positive values) and deficits in mm (negative) constitute one of the methods to assess visually periods of drought or water logging (file ccWsyyUA.*)
(7) The "rangeland-index" (RI; file ccWRyyOA.* or cc2RyyOA.*) is the classic FAO water satisfaction index computed for periods of 5 dekads, with average evapotranspiration kept at potential level (KCR=1) and an assumed WHC of 50 mm. Similar to moving averages, the value assigned to a dekad corresponds to the five-dekad period centred about that dekad. Thus 3 for dekads 1 to 5, 4 for dekads 2 to 6... The rangeland index, when plotted over time, assumes a smoothed aspect (Figure 1) and is associated with crop phenology. Based on local experience, there is usually a rather good association between planting and the RI. For instance, in Tanzania, it was found that the planting of millet usually occurs when the rising RI curve crosses the 0.4 (40%) threshold.
The RI is computed using actual rain (not "working rain" as described under 2.3, i.e. future values are coded as missing and not taken as their normal).
The dekads when the RI curve passes the 30 to 70% thresholds is given in the text output file ccWsyyOX.TXT or cc2syyOX.TXT .
