Assessment of anthropogenic impact on flood dynamics in Moldova using
hydrological modeling
Abstract
Main hydrological disasters that occur in the Republic of Moldova are
floods. In the last 70 years (1947-2015) floods caused losses of 583
mil. US$. Thereby, a proportion of 55% of the damage was determined by
flash floods and 40% - by fluvial floods. Increasing human impact on
environment is considered to be the main factor for modifications of
flood runoff regimes. Moldova features intense economic activity
resulting in the fact that over 60% of the territory is used for
agriculture and almost all rivers’ morphology is heavily modified.
Therefore, the evaluation of anthropogenic impact on floods generation
and propagation processes is of particular interest. Environmental
change scenario modelling was performed to analyze the modification of
flood wave hydrograph features with the help of the JAMS/J2000
hydrological model. 11 pilot basins have been utilized for this
analysis, which represent different parts of the country. Assessment of
land cover changes impact over the past 3 decades on flood runoff was
performed by consecutive change of land cover layer for 1982 and 2013
for each flood event. The results showed that land cover changes caused
a slight decrease of flood runoff in the northern part of the country
(-4-10%) due to processes of basins naturalization in areas that were
left fallow with undergoing natural succession. In the central and
southern rivers basins an increase of flood features could be identified
(+2-35%) because of intensified agricultural and urbanization
processes. Assessment of reservoir impact on floods was performed on
example of the Byc River, which is subject to change due to the
Ghidighici reservoir. Thus, first, hydrological model was calibrated and
validated for the Byc River natural runoff period and later applied for
runoff simulation for the reservoir operation period. Differences of
real and naturalized modeled runoff showed that Ghidighici reservoir
causes a significant decrease of flood runoff which is from 3 up to 20
times lower than in the case of natural runoff modeling. Resulted
hydrological models are well suited for land planning, hydraulic
engineering and flood control. They can serve as support for authorities
in decision making in flood protection and water resources management.