Geology Reference
In-Depth Information
Chapter 7
Data-Based Evapotranspiration Modeling
Abstract This chapter focuses on data-based modeling of evapotranspiration and
evaporation from three totally different eco-climatic regions. In the
first few sec-
tions, data-based modeling (arti
cial neural network) results are compared with
reference to evapotranspiration (ET
0
), estimated using traditional models from
meteorological data. The second section is fully dedicated to evaporation modeling
with data-based modeling concepts and input section procedures applied to evap-
oration modeling. In Sect.
7.1
, we describe the mathematical details of the reference
evapotranspiration models used. Analyses with traditional reference evapotranspi-
ration models are performed on data from the Brue catchment, UK and the Santa
Monica Station, USA. In Sect.
7.2
, studies are described which have been con-
ducted to see how data selection approaches respond to the evaporation data from
the Chahnimeh reservoirs region in Iran. In this case study, we consider compre-
hensive use of data selection approaches and machine learning AI approaches. We
have employed different model selection approaches such as GT, AIC, BIC,
entropy theory (ET), and traditional approaches such as data splitting and cross
correlation method on this daily evaporation data. Modeling with conventional
models and hybrid wavelet based models was performed as per recommendations.
7.1 Introduction
Evapotranspiration is very important in hydrological modeling as it represents a
substantial amount of moisture loss within the hydrological system. Several agri-
cultural and engineering disciplines rely on data-based evapotranspiration modeling
systems for better irrigation system design, hydrological modelling and
ood
warning, irrigation scheduling, and hydrologic and drainage studies. One can
nd
numerous models and methods in the literature for estimating evapotranspiration and
evaporation [
5
,
6
,
11
]. However, in recent decades, several computational techniques
and data-based statistical techniques have been successfully applied in the
field of
evaporation modeling using models such as ANN, Fuzzy, Neuro-fuzzy, SVM, and
