Premium
A hydrometeorological model for basin‐wide seasonal evapotranspiration
Author(s) -
Dias Nelson Luís,
Kan Akemi
Publication year - 1999
Publication title -
water resources research
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.863
H-Index - 217
eISSN - 1944-7973
pISSN - 0043-1397
DOI - 10.1029/1999wr900230
Subject(s) - evapotranspiration , hydrometeorology , streamflow , environmental science , precipitation , hydrology (agriculture) , surface runoff , potential evaporation , water storage , climatology , drainage basin , meteorology , geology , geography , cartography , ecology , geotechnical engineering , geomorphology , inlet , biology
A new methodology is proposed to capture the seasonal behavior of evapotranspiration from precipitation and streamflow data and to develop hydrometeorological evapotranspiration models tailored for each basin. The water budget method for determining evapotranspiration is downscaled to periods between 15 and 160 days that occur between well‐marked hydrological recessions. Using these uneven time periods, the error associated with the unknown soil moisture storage is minimized, whereas groundwater storage changes are estimated by means of a classical linear groundwater reservoir whose time constant is obtained by recession analysis. This seasonal water budget (SWB) method is able to reproduce the seasonal signal of evapotranspiration even when it is absent from the precipitation and streamflow records. The estimates are also compatible with calculated monthly net radiation. By selecting short enough water budget periods it is possible to check the relationship between SWB evapotranspiration estimates and net radiation, Penman and Priestley‐Taylor potential evaporation, precipitation minus outflow, water vapor deficit, and basin storage. The ratio of SWB evapotranspiration to an upper limit value represented by either net radiation or potential evaporation is well correlated with precipitation minus outflow, water vapor deficit, or both but is very poorly related to basin storage. The calculated regressions lead to a family of hydrometeorological evapotranspiration monthly (HEM) models fitted to the basins in question, in a way analogous to the calibration of rainfall‐runoff models. In the two watersheds where the methodology was applied the HEM models were able to preserve mass, with total accumulated differences no larger than 0.25 mm d −1 and root‐mean‐square errors of the order of 0.7 mm d −1 .