Abstract:
We analysed in the work how change in land use/land cover influences on flood characteristics
(frequency and magnitude) using a model inter-comparison approach, statistical
methods and two land use scenarios (land use scenario A and land use scenario B) for three
time horizons. The derived land use maps from these scenarios were considered as forcing
inputs to two physically based hydrological models (SWAT and WaSiM). The generalized
Pareto distribution combined with the Poisson distribution was used to compute flood frequency
and magnitude. Under land use scenario A, croplands increase at the annual rate
of 0.7% while under land use scenario B, it increases by 1.13% between 2003 and 2029.
The expansion of croplands indubitably enhances flood risks. Although there was a general
agreement about the sense of the variation, the magnitude of change in flood characteristics
was highly influenced by the model type. The rate of increase in flood quantiles simulated
from SWAT (0.36–1.3% for 10-year flood) was smaller than the corresponding magnitude
of changes simulated from WaSiM (2.6–7.0% for 10-year flood) whatever the scenarios.
The expansion of agricultural and pasture lands at the yearly rate of 0.7% under land use
scenario A (respectively, 1.13% under land use scenario B) leads to an increase of 3.6%
(respectively, 5.4%) in 10-year flood by considering WaSiM. This study is among the first
of its kind to establish a strong statistical relation between flood severity/frequency and
agricultural land expansion and natural vegetation reduction. The results of this study are
relevant and useful to the scientific research community as well as the decision makers for
framing appropriate policy decisions towards the management of extreme events and the
land use planning/management in future in the region.
