Help |
| CliFlo Home | close window |
The water balance scheme that is implemented in the "Climate Database" updates daily records in the rain table with estimates of soil moisture deficit and runoff. Runoff occurs when the deficit becomes negative and, at the other extreme, the deficit can become no greater than the available water capacity of the soil.
The deficit changes from day to day according to what rain fell and how much PET occurred. Any rain decreases the deficit and PET increases the deficit but for deficits greater than half the capacity of the soil the PET is linearly decreased by the proportion that the deficit is greater than half capacity. For example, if the deficit is 3/4 the capacity then only half the PET is added to the deficit or if the soil were empty then the effective PET would be reduced to zero.
CliDB calculations are based on field experiments using medium silt loam fully grassed soils, with available water capacity determined at 150 mm. Runoff is assumed to occur when the soil moisture content exceeds Field Capacity (typically about 38% by volume). The term runoff is taken to encompass both surface runoff and gravitational drainage through the soil profile.
However, an initial value of the soil moisture deficit may not be available so an initialising scheme is needed. The one adopted tracks the daily values for starts from both full and empty and when the deficit from the full start and that from the empty start have crossed or their difference is within 10% of the available water capacity, the mean of the 2 values is taken as the first correct value that can be stored. Initialisation also takes place whenever a significant number of daily rainfalls are missing. (Accumulations are shared out over the period of the accumulation and so those days are NOT missing).
Periods of missing rainfall data can be dealt with as just explained but it is, otherwise, essential that rainfall data be available. The PET data is preferentially the monthly Penman estimates at the station concerned but corrected Priestley-Taylor estimates are also allowed as are either from stations within a 50km radius and within +/- 200m of the station altitude. In this way all the monthly PETs for the period concerned might be found but a few missing months may remain and these can be filled with the particular monthly average (e.g. a JAN with the average of the available JANs).