Premium
AUTOMATIC CALIBRATION OF A HYDROLOGIC MODEL FOR SIMULATING GROUNDWATER TABLE FLUCTUATIONS ON FARMS IN THE EVERGLADES AGRICULTURAL AREA OF SOUTH FLORIDA
Author(s) -
Kwon Hoyoung,
Grunwald Sabine,
Beck Howard,
Jung Yunchul
Publication year - 2014
Publication title -
irrigation and drainage
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.421
H-Index - 38
eISSN - 1531-0361
pISSN - 1531-0353
DOI - 10.1002/ird.1822
Subject(s) - structural basin , hydrology (agriculture) , calibration , hydraulic conductivity , groundwater , environmental science , matching (statistics) , water quality , agriculture , water table , table (database) , drainage basin , mathematics , statistics , soil science , computer science , geography , geology , ecology , data mining , cartography , geomorphology , geotechnical engineering , biology , soil water
The Shuffled Complex Evolution—Universal Algorithm (SCE‐UA) is an automatic calibration algorithm that has shown success in finding a globally optimum objective function with more efficiency than other methods. We incorporated the SCE‐UA into our novel modeling environment, utilizing an ontology‐based simulation (OntoSim‐Sugarcane) framework adapted to analyze groundwater table (WT) fluctuations and drainage practices on four farm basins in the Everglades Agricultural Area of south Florida. Utilizing two water years (WY96–97) of farm WT fluctuations observed at a portion (<16 ha) of each farm basin, two parameters—lateral hydraulic conductivities of soil profile and vertical hydraulic conductivity of underlying limestone—were automatically calibrated. Regardless of farms, the best parameter sets that minimize the objective function of daily root mean square error could be found after 1500 simulation runs. The quality of matching simulated to observed values of farm WT were further assessed by the Nash–Sutcliffe efficiency coefficient (NSE). The NSE ranged from 0.38 to 0.75 (calibration period, WY96–97) and 0.10 to 0.76 (validation period, WY98–99) on all four farms. These results indicate that this coupling strengthens the capability of OntoSim‐Sugarcane to model hydrology by objectively finding the best parameter sets. Copyright © 2014 John Wiley & Sons, Ltd.