See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/364474134

Hydrological impacts of land use/cover changes in the Lake Victoria basin

Article in Ecological Indicators · December 2022

DOI: 10.1016/j.ecolind.2022.109580

CITATIONS READS

11 308

6 authors, including:

Yongwei Liu Guiping Wu

Chinese Academy of Sciences Nanjing Institute of Geography & Limnology, Chinese Academy of Sciences
31 PUBLICATIONS 415 CITATIONS 54 PUBLICATIONS 928 CITATIONS

SEE PROFILE SEE PROFILE

Xingwang Fan Guojing Gan

Chinese Academy of Sciences Nanjing Institute of Geography and Limmology, Chinese Academiy of Sciences
62 PUBLICATIONS 746 CITATIONS 30 PUBLICATIONS 435 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Yongwei Liu on 24 November 2022.

The user has requested enhancement of the downloaded file.

 Ecological Indicators 145 (2022) 109580

Contents lists available at ScienceDirect

Ecological Indicators
journal homepage: www.elsevier.com/locate/ecolind

Original Articles

Hydrological impacts of land use/cover changes in the Lake Victoria basin

Yongwei Liu a, Guiping Wu a, Xingwang Fan a, Guojing Gan a, Wen Wang b, Yuanbo Liu a, *
a
Key Laboratory of Watershed Geography Sciences, Nanjing Institute of Geography and Limnology, Chinese Academy of Sciences, Nanjing, China
b
State Key Laboratory of Hydrology-Water Resources and Hydraulic Engineering, Hohai University, Nanjing, China

A R T I C L E I N F O A B S T R A C T

Keywords: Understanding the hydrological impacts of land use and land cover (LULC) changes is significant for sustainable
LULC changes water resources management and planning. The Lake Victoria basin (LVB) has experienced extensive forest and
Lake Victoria basin grass degradation and agricultural land expansion under rapid socio-economic development and population
SWAT
growth in recent decades. However, the hydrological impacts of LULC changes for the whole LVB is still poorly
Drought
documented. This study first investigated the LULC changing effects of LVB, focusing on its impacts on the annual
and seasonal runoff and the hydrological drought based on a distributed hydrological modeling of the Soil and
Water Assessment Tool (SWAT). SWAT model showed comparatively good applicability in seasonal runoff
simulation of LVB, with the percent bias (PBIAS) kept within ± 20 % and the coefficient of determination (R2)
over 0.6 at ten hydrological stations available over the calibration and validation phases. Generally, the annual
runoff obtained a monthly increase of 1.5 mm under collective LULC changes. Moreover, the LULC effects
presented considerable seasonal dependence. A largest runoff increase of approximately 4 mm was detected in
short rainy season attributed to the combined surface runoff and groundwater increase. Insignificant runoff
increase was observed in long rainy and dry seasons under the complementary effects of surface runoff increase
and groundwater decrease. Additionally, the hydrological drought was generally aggravated with increased
drought frequency and lengthened duration, particularly for the central western and eastern regions with
massive conversion of forest to agricultural land. The findings provide importance implication for rational water
resources management and drought disaster response for the LVB.

1. Introduction resources in recent decades (Bosch and Hewlett, 1982; Stednick, 1996;
Woldeamlak and Sterk, 2005; Mango et al., 2011; Cuo, 2016; Gashaw
Land use and land cover (LULC) change is a significant driver of et al., 2018; Ahmed et al., 2022; Zhang et al., 2008; Cuo and Zhang,
global environmental change (Anand et al., 2018). As one of the most 2013; Chawla and Mujumdar, 2015; Yan et al., 2017; Rogger et al.,2017;
important human interventions to nature, changes in LULC mainly in­ Wang et al., 2018). The main LULC impacts investigated include
cludes the alteration of vegetation types (e.g. from forest, grass land to deforestation (afforestation), urbanization and conversion to agriculture
cropland) and the modification of land use practices. These changes tend (Remondi et al., 2016; Guimberteau et al., 2017; Braud et al., 2013;
to alter the hydrological and water cycle processes, imposing impacts on Lanckriet et al., 2012; Bonnesoeur et al., 2019; Jones et al., 2022).
the available water resources (both for surface runoff and groundwater) Studies have revealed that deforestation (afforestation) can cause an
and runoff regimes of catchments worldwide (Wagner et al., 2017; increase (decrease) in annual streamflow, but the effects are inconclu­
Gashaw et al., 2018; Javadi et al., 2021; Saatsaz et al., 2013). Under­ sive on seasonal scales and for peak and low flows (Cuo, 2016; Bruijn­
standing the hydrological impacts of LULC changes is imperative for zeel, 2004; Bonell et al., 2010). With an increased area of impervious
water resources planning and management (Vorosmarty et al., 2020; surfaces, urbanization tends to increase peak flow while reduce soil
Remondi et al., 2016). water infiltration (Miller et al., 2014; Remondi et al., 2016). The effects
LULC effects on hydrology and water resources have recently drawn of reclamation are largely dependent on the preceding land cover types
substantial attention from the hydrologic community, due to the (Cuo, 2016). Hydrological responses to LULC changes are complex
increasing human activities and rising scarcity of available water under various climate and geographical conditions due to complicated

* Corresponding author.
E-mail addresses: [email protected] (Y. Liu), [email protected] (G. Wu), [email protected] (X. Fan), [email protected] (G. Gan), [email protected]
(W. Wang), [email protected] (Y. Liu).

https://doi.org/10.1016/j.ecolind.2022.109580
Received 30 August 2022; Received in revised form 12 October 2022; Accepted 15 October 2022
Available online 19 October 2022
1470-160X/© 2022 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-
nc-nd/4.0/).
Y. Liu et al. Ecological Indicators 145 (2022) 109580

inter-relationships between various hydrological components such as Woldesenbet et al., 2017; Guzha et al., 2018). By using a streamflow
precipitation, evapotranspiration, infiltration, and runoff (Dwarakish model, Mutie at al. (2006) determined the extent and the effect of LULC
and Ganasri, 2015). LULC effects on streamflow or runoff are usually changes on the flow regime of the transboundary Mara River in Kenya
site-specific and are dependent on many factors including climate con­ and Tanzania. Githui et al. (2009) implemented the Soil and Water
dition, soil type, topography, spatial scale, and anthropogenic activities Assessment Tool (SWAT) model to investigate the LULC effects on the
(Chang et al. 2014; Wang et al. 2017; Zuo et al. 2016). A specific analysis runoff of River Nzoia catchment, Kenya. Baker and Miller (2013)
of hydrological responses to LULC changes is essential for sustainable adopted the SWAT model to assess the LULC effects on water resources
water resources utility and management. in the Njoro River watershed located in the southwestern portion of Rift
Africa experienced an extensive conversion from forest to agriculture Valley, Kenya. The above-mentioned researches mainly focused on small
and pasture over the past decades, secondary only to South America. basins within LVB. However, the hydrological impacts of the extensive
Nearly 13 million hectares of original forest have been lost and the LULC changes over the whole LVB are poorly documented. To quantify
remaining is fragmented and continually under threat in East Africa the LULC effects on hydrology in large basins (e.g. the whole LVB) still
(FAO, 2010). In particular for the Lake Victoria basin (LVB), the rapid remains a challenge, because where the interaction between LULC,
socio-economic development and growing population have brought climate characteristics and underlying hydrological processes are more
massive LULC changes, further leading to many environmental issues, complex and non-static than small basins (e.g. Uhlenbrook, 2004; van
such as droughts, floods, soil erosion, and land degradation (Tibebe and Griensven et al., 2012). In particular for entire LVB, the investigation
Bewket, 2011). Recently, some researchers have investigated the LULC endures more difficulty than other regions due to its data scarce con­
effects on runoff in LVB (e.g. Maitima et al., 2010; Mango et al., 2011; dition. Additionally, the existing researches mainly focused on the LULC

Fig. 1. (a) Location of Lake Victoria basin, (b) the elevation, major rivers, distribution of hydrological stations, and the delineated subbasins for SWAT streamflow
modeling, the multi-year monthly average (c) precipitation (Prep) and (d) temperature (Temp), and the intra-annual variations of (e) Prep and (f) Temp. Prep and
Temp are extracted from the CHIRPS and ERA5 dataset (Section 3.1.2).

2
Y. Liu et al. Ecological Indicators 145 (2022) 109580

effects on water resources or floods, the impacts on droughts remains through canopy interception, plan transpiration, and soil evaporation,
largely unexplored. both are sensitive to LULC changes.
Therefore, this study conducted a comprehensive investigation on
the hydrological impacts of LULC changes in the whole LVB over the 3.1.2. Model inputs
recent four decades based on a distributed hydrological modeling of the The input for SWAT model includes digital elevation model (DEM),
SWAT, with a focus on: i) the responses of runoff to LULC changes on LULC, soil, and climate data. In the current study, the DEM was derived
annual and seasonal scales; and ii) the impacts of LULC changes on the from the hydrological data and maps based on Shuttle Elevation De­
hydrological extremes of droughts. rivatives at multiple Scales with a spatial resolution of 15 arc-second
(http://hydrosheds.cr.usgs.gov). The LULC was extracted from the
2. Study area Regional Center for Mapping of Resources for Development (RCMRD)
with a 30-m resolution (https://geoportal.rcmrd.org/). To assess the
LVB is located in the upper reaches of the Nile River basin, East Af­ impact of LULC changes on hydrology, two LULC maps in 1985 and
rica (Fig. 1a). The whole basin occupies an area of approximately 2014 were utilized (Fig. 2a, b). The soil data were downloaded from the
251,000 km2, including the 68,900-km2 Lake Victoria (URT 2001). The Harmonized World Soil Database (HWSD) with a spatial resolution of 30
lake is surrounded by varied-sized mountains except in the north, with arc-second (https://www.fao.org/soils-portal/data-hub/soil-maps-and
an elevation from 937 to 4228 m (Fig. 1b). This terrain condition con­ -databases/harmonized-world-soil-database-v12/en/https://www.fao.
tributes to the drainage of ten more major rivers to Lake Victoria, e.g. the org/soils-portal/data-hub/soil-maps-and-databases/harmonized-world
river Kagera, Nozia, Simiyu, and Mara (Fig. 1b). LVB has an equatorial -soil-database-v12/en/https://www.fao.org/soils-portal/data-hub/soil-
hot and humid climate with the monthly precipitation varying from 50 maps-and-databases/harmonized-world-soil-database-v12/en/). The
to 200 mm and the temperature from 9 to 24 ℃ in space (Fig. 1c, d). soils for the whole catchment of LVB were classified into 21 categories
Intra-annually, LVB has a biannual rainfall pattern with a typical long (Fig. 2c). Climate data included the daily precipitation, maximum and
rainy season (March-May), a short rainy season (October-December), minimum temperature, wind speed, relative humidity, and solar radia­
and a dry season (June-September) (Fig. 1e, f). tion. The precipitation data (1981–2020) were collected from the 0.05
degree Climate Hazards Group Infrared Precipitation with Station Data
3. Methods and materials (CHIRPS) dataset (https://coastwatch.pfeg.noaa.gov/erddap/gridda
p/chirps20GlobalDailyP05.html). The other climate data were extrac­
3.1. Hydrological modeling ted from the fifth generation (ERA5) of the European Centre for
Medium-Range Weather Forecasts (ECMWF) reanalysis dataset for
3.1.1. The SWAT model global climate and weather (https://cds.climate.copernicus.eu/cdsa
SWAT is a physically based, spatially distributed, and temporally pp#!/dataset/reanalysis-era5-single-levels?tab=form).
continuous hydrological model (Neitsch et al., 2011). It was designed to
predict the impacts of land management practices on water, sediment, 3.1.3. Model calibration and validation
and agricultural chemical yields in large and complex catchments The SWAT model for LVB was built up based on the underlying
(Arnold et al., 2012), and provides flexibility for modification to capture surface and climate data (section 3.1.2). The catchment was divided into
different components of hydrological processes in a given catchment a total of 89 subbasins (Fig. 1b). To simplify and improve model cali­
(Jafari et al., 2021; Einia et al., 2020). SWAT also runs with minimum bration, a sensitivity analysis of model parameters to rainfall-runoff
data inputs, which makes it advantageous in regions with limited data. modeling was performed before model calibration using the Latin-
The applicability of SWAT model in poorly-gauged rural African hypercube and one-factor-at-a-time method (McKay et al., 1979; Mor­
catchments has been approved, given appropriate satellite-based rainfall ris, 1991). The first ten sensitive model parameters were listed in
estimates and proper manual or automatic calibration (Mango et al., Table 1. Sensitive model parameters were adjusted based on measured
2011; Gashaw et al., 2018). discharges (after reduction treatment) from the 10 hydrological stations
In SWAT modeling, the catchment was geographically divided into in LVB over the period of 1981–1987 (Fig. 1b). The reduction treatment
subbasins, which were further delineated into several hydrological for the measured discharges was based on a global gridded monthly
response units (HRUs) with the same land cover types, soil properties, water withdrawal product developed by Huang et al. (2018). This pro­
and terrains (Liu et al., 2017). Therefore, SWAT can assess and predict cessing is to take the impacts of water withdrawal of human beings into
the hydrological impacts of LULC changes on subbasins or HRUs. The account. In the above parameter adjustment, the Sequential Uncertainty
hydrological cycle simulated by SWAT is based on the following water Fitting version 2 (SUFI-2) algorithm (Abbaspour et al., 2004; 2007) and
balance equation: the manual tuning were applied in combination. The model parameters
for the whole catchment were classified based on the same landuse, soil
∑
t
types, terrain or geological condition. The classified parameters were
SW t = SW 0 + (Pday − Qsurf − Ea − wseep − Qgw )
i=1
calibrated based on the streamflow measurements at the downstream
observation station or the station in the adjacent river basins. The model
where SWt and SW0 are the final and initial soil water content, t is the validation was based on the measured discharges (after reduction
time (days), Pday is the amount of precipitation on day i, Qsurf is the treatment) from the 10 hydrological stations in the period of 1988–1991
amount of surface runoff, Ea is the amount of evapotranspiration, wseep is with available data. The model performance was evaluated visually and
the amount of water entering the vadose zone from the soil profile, and by using the percent bias (PBIAS) and coefficient of determination (R2)
Qgw is the amount of return flow. on seasonal scales (van Griensven et al., 2012). The LULC map of 1985
The surface runoff simulation is generally based on the Soil Con­ was utilized for model calibration and validation, considering the time
servation Service (SCS) curve number method (SCS, 1972), which pro­ of streamflow measurements available. To assess the hydrological im­
vides consistent basis for estimating the amounts of runoff under varying pacts of the historical LULC changes, the calibrated SWAT model was
land use and soil types (Rallison and Miller, 1981). After surface runoff run for LULC maps of 1985 and 2014, respectively, and the hydrological
generation, the water infiltrated or percolated to soil profile is redis­ conditions were compared.
tributed using a storage routing technique with the field capacity of soil
as the threshold. The percolation generated from the bottom of the soil 3.2. Drought identification and characterization
profile is the source of groundwater recharge. During the above surface
and groundwater generation processes, large amounts of water is lost The hydrological extremes of droughts were identified by using the

3
Y. Liu et al. Ecological Indicators 145 (2022) 109580

Fig. 2. LULC maps of Lake Victoria basin for (a) 1985 and (b) 2014 from the Regional Center for Mapping of Resources for Development, and (c) soil map from the
Harmonized World Soil Database.

of one drought event and characterized by the quotient of the drought

Table 1
severity and drought duration (Liu et al., 2021).
The first ten sensitive model parameters in SWAT for Lake Victoria basin.
Sequence Parameter Description 4. Results
name

1 CN2 moisture condition II curve number 4.1. LULC changes over recent decades
2 aqshthr,q threshold water level in shallow aquifer for
baseflow
3 AWCly available water capacity of soil at layer ly LVB has experienced significant LULC changes forced by high pop­
4 βrev revap coefficient ulation growth, poor land management, and unsustainable land uses
5 z depth from soil surface to bottom of soil layer over the recent decades. Two phases of LULC maps for SWAT modeling
6 αgw baseflow recession constant (reclassification based on the RCMRD land use data) show that a sub­
7 n Manning’s value for overland flow
8 canmx maximum canopy storage
stantial portion of forest and grass land was converted to agricultural
9 epco plant uptake compensation factor land over 1985–2014 (Fig. 3). The agricultural land expansion seems to
10 δgw delay time for aquifer recharge be ubiquitous in the whole catchment. The urban areas also expanded
rapidly, especially in the western part of LVB (Fig. 3). Statistics for the
89 subbasins in SWAT modeling clearly showed the LULC changes
standardized runoff index (SRI; Shukla and wood, 2008) based on the
(Fig. 4). The evergreen (mixed) forest area decreased from 5 % (20 %) to
model simulated runoff data. The calculation of SRI is similar to that of
1 % (10 %) from 1985 to 2014. The grassland area decreased from 30 %
the standardized precipitation index (SPI; McKee et al., 1993), with
in 1985 to 10 % in 2014. The wetland area also showed a substantial
merely difference in their inputs. We focused on the 3-month SRI to
decline. However, the agricultural area increased from approximately
account for seasonal droughts in the catchment. The SRI less than –1 was
35 % in 1985 to 65 % in 2014. The urban area also increased yet not
regarded as being under drought as the moderately to extremely dry
clearly shown at this scale due to its little percentages in the whole basin.
conditions were concerned (Guttman, 1999). The drought events were
Note that considerable differences exist among subbasins (Fig. 4).
identified based on the SRI time series of each subbasin using the theory
of run (Yevjevich, 1967). To reduce the perturbation of the outliers on
drought event identification, two adjacent droughts were pooled to form 4.2. Performance of the SWAT in runoff modeling
a new drought when the time interval was less than 1 month (Wang
et al., 2021). Fig. 5 compares the SWAT model simulated with the observed
The following drought characteristics were accounted for in the streamflow at ten hydrological stations available in this study. The
aforementioned drought event identification framework: (1) drought hydrographs of both streamflow data generally match well for all sta­
duration: time span between the initiation and termination of one tions during the calibration and validation phases. The statistics of the
drought event; (2) drought severity: indicator of the degree of water model performance in runoff simulation also illustrate the consistency of
deficiency and characterized by the cumulative SRI for one drought the simulations and observations (Table 2). For model calibration, the
event duration; (3) drought intensity: indicator of the average strength percent bias (PBIAS) values are kept within ±20 % and the coefficient of
determination (R2) values exceed 0.6 for all stations except for SIMIYU,

4
Y. Liu et al. Ecological Indicators 145 (2022) 109580

Fig. 3. LULC maps in (a) 1985 and (b) 2014 for SWAT modeling. AGRR, FRSE, FRST, HAY, RNGE, URMD, WATR, and WETL are agricultural land-row crops,
evergreen forest, mixed forest, hay, range-grasses, residential-medium density, water, and wetlands, respectively.

4.3. The impacts of LULC changes on runoff

To evaluate the hydrological impacts of the massive conversion of

forest and grass to agricultural land in LVB, the SWAT model simulated
runoff (streamflow) based on the LULC maps of 1985 and 2014 were
compared. Considering the potential difference of the intra-annual
climate conditions on the LULC effects on hydrology, the typical long
rainy season (Mar–May), the short rainy season (Oct–Dec), the dry
season (Jun–Sep), and the transitional season (Jan–Feb) were investi­
gated independently. Fig. 6 shows the difference of the catchment
Fig. 4. Percentages (%) of subbasin LULC types in 1985 (left) and 2014 (right).
average runoff depths between the LULC of 2014 and 1985 based on the
HAY, URMD, WETL, WATR, AGRR, FRSE, RNGE, and FRST are hay, residential-
statistics of forty years from 1981 to 2020. Generally, the runoff gen­
medium density, wetlands, water, agricultural land-row crops, evergreen forest,
range-grasses, and mixed forest, respectively.
eration increases due to forest and grass land degradation and farm land
expansion. The monthly runoff depth increases by approximately 1.5
mm for the whole catchment. Intra-annually, the runoff increases
MURONGWE and RUSUMO. For model validation, the PBIAS values are
exhibit considerable differences, with larger increases (~4 mm) in the
kept within ±20 % except for RWAMBWA, and the R2 values are over
short rainy season than others. The runoff increases in long rainy season
0.7 for most stations. The relatively high consistent hydrographs of the
and drought season contains some uncertainty as the runoff decrease
simulated and observed streamflow and the satisfying model perfor­
exists in individual years. Additionally, the inter-annual climate change
mance statistics indicate good applicability of the constructed model in
in the forty years also complicate the hydrological responses to LULC
streamflow modeling in LVB.
changes, in particular for the short rainy season.

Fig. 5. Comparison of the SWAT model simulated streamflow and the measured streamflow during calibration (1981–1987) and validation (1988–1991) phases at
ten hydrologic stations (Fig. 1b).

5
Y. Liu et al. Ecological Indicators 145 (2022) 109580

Table 2 the western perimeter and the eastern lake shore (Fig. 7a). The large
Statistics of the SWAT model simulated streamflow compared with the measured areas of yearly runoff increases might have benefited from the runoff
streamflow at ten hydrologic stations. increase in short rainy and transitional seasons, as the spatial patterns
Hydrological Stations/Rivers Calibration Validation for their runoff differences are similar (Fig. 7a, b, e). For the long rainy
(1981–1987) (1988–1991) and dry seasons, the runoff increases caused by LULC changes are not
PBIAS (%) R2 PBIAS (%) R2 robust (Fig. 7c, d). The runoff decrease exists in part regions of north­
west, southwest, northeast, and southeast.
SIMIYU 6.40 0.44 − 8.78 0.80
BURASIRA 2.84 0.60 8.65 0.86
MURONGWE − 6.58 0.54 − 7.63 0.88
KIBAYA − 7.20 0.82 − 14.25 0.77
4.4. The impacts of LULC changes on droughts
NGOZI-BUTARE 0.35 0.78 − 13.11 0.81
KIGALI − 2.57 0.71 Fig. 8 shows the differences of the SRI based on the LULC maps of
RUSUMO − 9.18 0.54 2014 and 1985 for the whole catchment. The wide range of SRI differ­
RWAMBWA 5.34 0.69 − 29.22 0.95
ences (positive and negative) indicates a clear impact of LULC changes
KAGERA 16.62 0.77 − 8.17 0.62
MARA 20.96 0.71 on hydrological droughts in LVB. Considering the spatial heterogeneity
of droughts and the substantial spatial differences of LULC changes on
runoff, drought events were investigated for each subbasin. The statis­
tics for the drought characteristics of the events identified on total 89
subbasins based on the LULC of 1985 and 2014 are presented and
compared (Fig. 9). Generally, it can be observed that the drought fre­
quency (indicated by the number of drought events) has a considerable
increase due to LULC changes (Fig. 9a). The drought density is slightly
enhanced (Fig. 9d). Nevertheless, the drought duration and severity do
not present significant changes in statistics of the 89 subbasins (Fig. 9b,
c).
The spatial distribution of the effects of LULC changes on drought
characteristics is presented in Fig. 10. The effects are qualified as the
differences in the number of identified events, the long-term average
drought duration, severity and intensity between the LULC of 2014 and
1985. Spatially, the drought frequency dominantly increases in the
Fig. 6. The difference (Differ) of the catchment average runoff based on the
LULC of 2014 and 1985 for the whole year (Year), long (Mar–May) and short
rainy (Oct–Dec) season, dry season (Jun–Sep), and the transitional season
(Jan–Feb) based on the statistics of forty years over 1981–2020. Jan, Feb, Mar,
Jun, Sep, Oct, Dec are abbreviated for January, February, March, June,
September, October, and December, respectively. mon represents month.

Runoff responses to LULC changes are diverse in space. Fig. 7 shows

the multi-year average runoff differences between the LULC of 2014 and
1985 for the whole year and the abovementioned four seasons at sub­ Fig. 8. Difference (Differ) of the standard runoff index (SRI) between the LULC
basin level. For the whole year, runoff increased in most regions except of 2014 and 1985 for the whole Lake Victoria basin.

Fig. 7. Maps of the multi-year average runoff difference (dt.runoff) between the LULC of 2014 and 1985 for (a) the whole year (Year), (b) the transitional period
(Jan–Feb), (c) the long rainy season (Mar–May), (d) the dry season (Jun–Sep), and (e) the short rainy season (Oct–Dec). Jan, Feb, Mar, Jun, Sep, Oct, Dec are
abbreviated for January, February, March, June, September, October, and December, respectively.

6
Y. Liu et al. Ecological Indicators 145 (2022) 109580

Fig. 9. Probability distribution of the (a) number of drought events, the average (b) drought duration, (c) drought severity, and (d) drought intensity for the
identified events over 1981–2020 in the total 89 LVB subbasins for the LULC of 1985 and 2014 (Fig. 1b). The vertical dashed lines indicate the value with the highest
probability density.

Fig. 10. Maps of the differences in the (a) number of drought events, the average (b) drought duration, (c) drought severity, and (d) drought intensity for the drought
events identified over 1981–2014 between the LULC of 2014 and 1985 for each subbasin (Fig. 1b).

7
Y. Liu et al. Ecological Indicators 145 (2022) 109580

central west and southeast, while decreases in the western perimeter and groundwater in short rainy season strengthens the runoff increase
the north and central east of the LVB (Fig. 10a). The average drought dominated by the surface runoff. Guzha et al. (2018) also showed that
duration increases substantially in the central west and east, while forest cover loss increased annual stream discharges and surface runoff,
slightly decreased in other regions (Fig. 10b). The impacts of LULC while decreased low flows in field studies in East Africa. Baker and
changes on drought severity are in consistent with those on drought Miller (2013) noted that the streamflow has increased and groundwater
duration. Drought becomes aggravated in the central west and east, recharge has decreased due to change in land use in the Nioro basin
shown as long term average drought severity (Fig. 10c). Regarding using SWAT model. Other studies have also reported the increase of
drought intensity, the effects caused by LULC changes are slight for most surface runoff in regions of the LVB due to the expansion of cultivated
regions of LVB as where the difference statistics remain within ± 0.1. lands and reduction of vegetation covers, e.g. Karamage et al. (2017) in
The increased drought conditions are predominantly enforced by the Rwanda, Githui et al. (2009) in Nzoia catchment in Kenya.
LULC effects on drought frequency and duration. The impacts of LULC changes on runoff are embodied through
changes in canopy interception, root water uptake, and infiltration (Cuo,
5. Discussion 2016). Deforestation would decrease the leaf area, which consequently
reduces canopy interception rate and plant transpiration, and thus in­
5.1. Reasonability of the increased runoff caused by LULC changes creases the surface runoff generation (Woldesenbet et al., 2017; Anand
et al., 2018). Transforming forest to agricultural land probably domi­
The total runoff of LVB is demonstrated to be increased due to the nates the hydrological responses to LULC changes in dry and short rainy
substantial decrease of forest and grass land and increase of agricultural seasons over LVB (Fig. 11a, c). For the long rainy and transitional sea­
and urban land. Moreover, the runoff increases present seasonality. sons, the hydrological impacts of the conversion from grassland to
Regarding the source of total runoff, it consists of the surface runoff and agricultural land cannot be ignored as in the meantime the actual
groundwater or baseflow. Generally, the annual and seasonal runoff evapotranspiration is increased (Fig. 11c). The increased evapotranspi­
increases for the whole catchment should be dominated by the surface ration might be related to the fact that the rising water demand for crops
runoff as the annual groundwater is slightly decreased (Fig. 11a,b). The during growing season with high temperature (Fig. 1f) likely exceeds
groundwater only contributes to the variation of the runoff over long that of the range-grass and counteracts the evapotranspiration decrease
and short rainy seasons. Specifically, the decreased groundwater in long caused by deforestation. Groundwater variations are largely influenced
rainy season weakens the magnitude of runoff increase and brings more by infiltration (percolation), which is complementary to changes in
uncertainties to the runoff change (see Fig. 6). The increased evapotranspiration (Fig. 11c, d). The opposite hydrological impacts of

Fig. 11. Catchment average (a) surface runoff (surfQ), (b) groundwater or baseflow (gwQ), (c) actual evapotranspiration (aET), and (d) water percolation
(Percolation) based on the LULC of 1985 (left) and 2014 (right) for the whole year (Year), long (Mar–May) and short rainy (Oct–Dec) season, dry season (Jun–Sep),
and the transitional season (Jan–Feb) based on the statistics of forty years over 1981–2020. Jan, Feb, Mar, Jun, Sep, Oct, Dec are abbreviated for January, February,
March, June, September, October, and December, respectively. mon represents month.

8
Y. Liu et al. Ecological Indicators 145 (2022) 109580

LULC changes on groundwater in long and short rainy seasons (Fig. 11b) The degradation in storage capacity can be largely attributed to the
should be related to the different climate conditions and vegetation substantial deforestation over 1985–2014, which is more evident in
growing status. The long rainy season is generally characterized by central western and eastern LVB (see Fig. 3). Deforestation would reduce
higher precipitation rate and temperature, and more vigorous growing the canopy interception and accelerate the runoff generation process
crops than short rainy season (Lake Victoria Basin Commission and (Lima et al., 2014; Lawrence and Vandecar, 2015). The lengthened
GRID-Arendal. 2017). The more vigorous crops would demand more drought duration might be related to the increase in evapotranspiration
water. Thus, it might weaken the evapotranspiration decreasing impact mainly caused by the vigorous crops in their growing season (Fig. 11c).
caused by transforming forest to agriculture land, while strengthen the
evapotranspiration increasing impact caused by transforming grassland 6. Conclusion
to agricultural land. Therefore, different hydrological impacts of LULC
changes on groundwater in long and short rainy seasons tend to occur. In This study investigated the hydrological impacts of LULC changes in
addition, the increase in impervious surface areas through urbanization the LVB over 1985–2014 based on a distributed hydrological modeling
is inclined to cause the increase in surface runoff, decrease in infiltra­ of the SWAT. The investigation focused on the LULC impacts on annual
tion, and then decline in groundwater, in particular for long rainy season and seasonal runoff and the hydrological extremes of droughts. The
with high precipitation rate (Olivera and DeFee, 2007; Remondi et al., main conclusions are as follows:
2016). (i) The LVB experienced extensive LULC changes over 1985–2014.
The specific LULC changes present substantial differences among sub­
5.2. Interpretation on the impacts of LULC changes on droughts basins. Generally, the evergreen and mixed forest decreased from 5 %
and 20 % to 1 % and 10 %, respectively; the grassland decreased from
Drought is one of the most severe natural disasters in the LVB 30 % to 10 %; the agricultural land increased from approximately 35 %
(Awange, 2007, 2021). Drought disasters threaten food security and to 65 %.
could lead to a catastrophe, since a majority of the population in the LVB (ii) The SWAT model presents relatively good applicability in sea­
are poor and living below poverty line (Skees, 2000; Agola and Awange, sonal runoff modeling in the LVB. The hydrographs of the simulated and
2014). Recent studies have shown some worrying scenes of drought observed streamflow have comparatively high consistency. The PBIAS
patterns attributed to natural climate change and increasing human values are basically within ± 20 % and the R2 values are over 0.6 for ten
influence (Awange et al., 2008, 2013; Swenson and Wahr, 2009). LULC hydrological stations available in both calibration and validation
changes are regarded as a critical human intervention to nature. The phases.
LVB experienced large-scale and significant deforestation, grass (iii) The annual runoff generally increased under the collective LULC
removal, and rapid expansion of agriculture and urbanization forced by changes of deforestation, grass removal, and agriculture expansion over
high population growth and rapid socio-economic development in 1985–2014. Multi-year monthly runoff depth increases approximately
recent decades (Section 4.1). We demonstrated that the abovementioned 1.5 mm for the whole catchment. The impacts of LULC changes on runoff
LULC changes generally aggravated the hydrological drought condition present considerable seasonal differences. The largest runoff increase of
in the LVB. The aggravated droughts are mainly manifested by the 4.0 mm was observed in short rainy season from the combined surface
increased drought frequency and the lengthened drought duration, in runoff and groundwater increase. Insignificant runoff increase was
particular for the central western and eastern LVB (Section 4.4). The detected in long and dry season due to the complementary effects of
results were supported by Mango et al. (2011), who demonstrated that surface runoff increase and groundwater decrease. Additionally, sub­
the conversion of forests to agriculture and grass land in Mara river basin stantial spatial heterogeneity exists in the LULC effects. The annual
in central eastern LVB is likely to reduce dry season flows. The increase runoff increase occurs in most regions except for the western perimeter
in drought (indicated by the total runoff) frequency is probably and central eastern lake shore. Similar spatial patterns are observed in
contributed by the surface runoff. Because a non-ignorable increment in short rainy and transitional seasons. Nevertheless, for long rainy and dry
the dispersion degree (indicated by the standard deviation) of the sur­ seasons, the runoff decrease occurs in part regions of northwest,
face runoff is observed, while it is not found in the groundwater southwest, northeast, and southeast.
(Fig. 12a, b). The increased dispersion of the surface runoff indicates the (iv) The hydrological drought condition of the catchment is generally
degradation of the catchment in its storage capacity on rainfall-runoff. aggravated by the LULC changes over 1985–2014, which is manifested

Fig. 12. Probability distribution of the (a) standard deviation of surface runoff (Std.of surfQ) and (b) groundwater or baseflow (Std.of gwQ) for the yearly runoff of
1981–2020 in the 89 subbasins of the LVB based on the LULC of 1985 and 2014 (Fig. 1b). The vertical dashed lines indicate the value with the highest probabil­
ity density.

9
Y. Liu et al. Ecological Indicators 145 (2022) 109580

in the increased drought frequency and the lengthened drought dura­ Awange, J.L., Forootan, E., Kusche, J., Kiema, J.K.B., Omondi, P., Heck, B., Fleming, K.,
Ohanya, S., Gonçalves, R.M., 2013. Understanding the decline of water storage
tion, in particular for the central western and eastern LVB.
across the Ramser-lake Naivasha using satellite-based methods. Adv. Water Res. 60,
Overall, the current study provides a comprehensive investigation on 7–23. https://doi.org/10.1016/j.advwatres.2013.07.002.
the hydrological impacts of the LULC changes for the whole catchment Baker, T.J., Miller, S.N., 2013. Using the Soil and Water Assessment Tool (SWAT) to
of the Lake Victoria. The reasonability of the runoff responses both for assess land use impact on water resources in an East African watershed. J. Hydrol.
486, 100–111.
the normal and drought status was also interpreted from the water cycle Bonell, M., Purandara, B.K., Venkatesh, B., Krishnaswamy, J., Acharya, H.A.K., Singh, U.
and rainfall-runoff process. This innovative work could deepen our V., Jayakumar, R., Chappell, N., 2010. The impact of forest use and reforestation on
knowledge on the impact of LULC changes on hydrology in the LVB. The soil hydraulic conductivity in the Western Ghats of India: implications for surface
and sub-surface hydrology. J. Hydrol. 391, 47–62. https://doi.org/10.1016/j.
quantitative analysis could serve as an important implication for the jhydrol.2010.07.004.
stakeholder and decision makers in water resources planning and Bonnesoeur, V., Locatelli, B., Guariguata, M.R., Ochoa-Tocachi, B.F., Vanacker, V.,
management. Additionally, the neglected feedbacks of the LULC Mao, Z., Stokes, A., Mathez-Stiefel, S., 2019. Impacts of forests and forestation on
hydrological services in the Andes: a systematic review. Forest Ecol. Manag. 433,
changes (e.g. deforestation) on climate may introduce certain uncer­ 569–584.
tainty to the hydrological impacts, which will be the topic of our future Bosch, J.M., Hewlett, J.D., 1982. A review of catchment experiments to determine the
work. effect of vegetation changes on water yield and evapotranspiration. J. Hydrol. 55,
3–23.
Braud, I.P., Breil, F., Thollet, M., Lagouy, F., Branger, C., Jacqueminet, S.K., Michel, K.,
2013. Evidence of the impact of urbanization on the hydrological regime of a
Declaration of Competing Interest
medium-sized periurban catchment in France. J. Hydrol. 485, 5–23.
Bruijnzeel, L.A., 2004. Hydrological functions of tropical forests: not seeing the soil for
The authors declare that they have no known competing financial the trees? Agric. Ecosyst. Environ. 104, 185–228. https://doi.org/10.1016/j.
interests or personal relationships that could have appeared to influence agee.2004.01.015.
Chang, H., Johnson, G., Hinkley, T., Jung, I.-I.-W., 2014. Spatial analysis of annual runoff
the work reported in this paper. ratios and their variability across the contiguous U.S. J. Hydrol. 511, 387–402.
https://doi.org/10.1016/j.jhydrol.2014.01.066.
Data availability Chawla, I., Mujumdar, P.P., 2015. Isolating the impacts of land use and climate change
on streamflow. Hydrol. Earth Syst. Sci. 19, 3633–3651.
Cuo, L., 2016. In: Tang, Q., Oki, T. (Eds.), Land Use/cover Change Impacts on Hydrology
Data will be made available on request. in Large River Basins: A Review Terrestrial Water Cycle and
ClimateChange—Natural and Human-Induced Impacts. American Geophysical
Union, Wiley, Washington DC, USA, pp. 103–134.
Acknowledgement Cuo, L., Zhang, Y., 2013. The impacts of climate change and land cover/use transition on
the hydrology in the Upper Yellow River Basin. China. J. Hydro. 502, 37–52. https://
doi.org/10.1016/j.jhydrol.2013.08.003.
This work was supported by the National Key R&D Program of Dwarakish, G.S., Ganasri, B.P., 2015. Impact of land use change onhydrological systems:
China, No.2018YFE0105900, the National Natural Science Foundation A review of current modeling approaches, Cogent Geoscience, 1:1, 1115691. doi:
10.1080/23312041.2015.1115691.
of China, No.41901049, No.41971042.
Einia, M.R., Javadia, S., Delavarb, M., Gassmanc, P.W., Jarihani, B., 2020. Development
of alternative SWAT-based models for simulating water budget components and
streamflow for a karstic-influenced watershed. Catena 195, 104801.
Data availability FAO – Food and Agriculture Organization of the United Nations:Global Forest Resources
Assessment Main report, FAO Forestry Paper 163, Food and Agriculture
The Dem data is available at http://hydrosheds.cr.usgs.gov. The Organization of the United Nations, Rome, 2010.
Gashaw, T., Tulu, T., Argaw, M., Worqlul, A.W., 2018. Modeling the hydrological
LULC data of RCMRD can be drawn from http://geoportal.rcmrd.org/. impacts of land use/land cover changes in the Andassa watershed, Blue Nile Basin,
The soil data of HWSD is from https://www.fao.org/soils-portal/data Ethiopia. Sci. Total Environ. 619–620, 1394–1408. https://doi.org/10.1016/j.
-hub/soil-maps-and-databases/harmonized-world-soil-database-v12/e scitotenv.2017.11.191.
Githui, F., Mutua, F., Bauwens, W., 2009. Estimating the impacts of land-cover change on
n/. The CHIRPS precipitation data can be collected from https://coast­ runoff using the soil and water assessment tool (SWAT): case study of Nzoia
watch.pfeg.noaa.gov/erddap/griddap/chirps20GlobalDailyP05.html. catchment, Kenya. Hydrolog. Sci. J. 54 (5), 899–908.
The ERA5 climate data is from https://cds.climate.copernicus.eu/cdsa Guimberteau, M., Ciais, P., Ducharne, A., Boisier, J.P., Dutra Aguiar, A.P., Biemans, H.,
De Deurwaerder, H., Galbraith, D., Kruijt, B., Langerwisch, F., Poveda, G.,
pp#!/dataset/reanalysis-era5-single-levels?tab=form.
Rammig, A., Rodriguez, D.A., Tejada, G., Thonicke, K., Von Randow, C., Von
Randow, R.C.S., Zhang, K., Verbeeck, H., 2017. Impacts of future deforestation and
References climate change on the hydrology of the Amazon Basin: a multi-model analysis with a
new set of land-cover change scenarios. Hydrol. Earth Syst. Sci. 21, 1455–1475.
https://doi.org/10.5194/hess-21-1455-2017.
Abbaspour, K.C., Johnson, C.A., van Genuchten, M.T., 2004. Estimating uncertain flow
Guttman, N.B., 1999. Accepting the standardized precipitation index: a calculation
and transport parameters using a sequential uncertainty fitting procedure.
algorithm. J. Am. Water Resour. As. 35, 311–322. https://doi.org/10.1111/j.1752-
J. Vadouse Zone 3 (4), 1340–1352.
1688.1999.tb03592.x.
Abbaspour, K.C., Yang, J., Maximov, I., Siber, R., Bogner, K., Mieleitner, J., Zobrist, J.,
Guzha, A.C.I., Rufino, M.C., Okoth, S., Jacobs, S., Nóbrega, R.L.B., 2018. Impacts of land
Srinivasan, R., 2007. Modelling hydrology and water quality in the pre-alpine/alpine
use and land cover change on surface runoff, discharge and low flows: Evidence from
Thur watershed using SWAT. J. Hydrol. 333, 413–430.
East Africa. J. Hydrol.:Regional Studies 15, 49–67.
Agola, N.O., Awange, J.L., 2014. Globalized poverty and environment. 21st Century
Huang, Z., Hejazi, M., Li, X., Tang, Q., Vernon, C., Leng, C., Liu, Y., Döll, P., Eisner, S.,
challenges and innovative solutions. Springer, Berlin.
Gerten, D., Hanasaki, N., Wada, Y., 2018. Reconstruction of global gridded mnthly
Ahmed, N., Wang, G., Booij, M.J., Sun, X., Hussain, F., Nabi, G., 2022. Separation of the
sectoral water withdrawals for 1971–2010 and analysis of their spatiotemporal
Impact of Landuse/Landcover Changeand Climate Change on Runoff in the
patterns. Hydrol. Earth Syst. Sci. 22, 2117–2133. https://doi.org/10.5194/hess-22-
Upstream Areaof the Yangtze River, China. Water Recour. Manag. 36, 181–201.
2117-2018.
Anand, J., Gosain, A.K., Khosa, R., 2018. Prediction of land use changes based on Land
Jafari, T., Kiem, A.S., Javadi, S., Nakamura, T., Nishida, K., 2021. 2021, Fully integrated
Change Modeler and attribution of changes in the water balance of Ganga basin to
numerical simulation of surface water-groundwater interactions using SWAT-
land use change using the SWAT model. Sci. Total Environ. 644, 503–519. https://
MODFLOW with an improved calibration tool. J. Hydrol.-Reg. Stud. 35, 100822.
doi.org/10.1016/j.scitotenv.2018.07.017.
Javadi, S., Saatsaz, M., Shahdany, S.M.H., Neshat, A., Milan, S.G., Akbari, S., 2021.
Arnold, J., Kiniry, J., Srinivasan, R., Williams, J., Haney, E., Neitsch, S., 2012. Soil and
A new hybrid framework of site selection for groundwater recharge. Geosci. Front.
Water Assessment Tool. Input/Output Documentation Version 2012. Texas Water
12, 101144.
Resources Institute TR-439. 654 pp.
Jones, J., Ellison, D., Ferraz, S., Lara, A., Wei, X., Zhang, Z., 2022. Forest restoration and
Awange, J., 2021. Lake Victoria Basin: droughts and food security. In: Lake Victoria
hydrology. Forest Ecol. Manag. 520, 120342.
Monitored From Space. Springer, Cham. https://doi.org/10.1007/978-3-030-60551-
Karamage, F., Zhang, C., Fang, X., Liu, T., Ndayisaba, F., Nahayo, L., Kayiranga, A.,
3_12.
Nsengiyumva, J., 2017. Modeling rainfall-runoff response to land use and land cover
Awange, J.L., Aluoch, J., Ogallo, L., Omulo, M., Omondi, P., 2007. Frequency and
changes in Rwanda (1990–2016). Water 9, 147.
severity of drought in the Lake Victoria region (Kenya) and its effects on food
Lake Victoria Basin Commission and GRID-Arendal. 2017. Lake Victoria Basin: Atlas of
security. Clim. Res. 33, 135–142. https://doi.org/10.3354/cr033135.
Our Changing Environment. Lake Victoria Basin Commission and GRID-Arendal,
Awange, J.L., Ogallo, L., Kwang-Ho, B., Were, P., Omondi, P., Omute, P., Omulo, M.,
Kisumu and Arendal.
2008. Falling Lake Victoria water levels: is climate a contribution factor? Clim.
Change. 89, 281–297. https://doi.org/10.1007/s10584-008-9409-x.

10
Y. Liu et al. Ecological Indicators 145 (2022) 109580

Lanckriet, S., Araya, T., Cornelis, W., Verfaillie, E., Poesen, J., Govaerts, B., Bauer, H., Saatsaz, M., Sulaiman, W.N.A., Eslamian, S., Javadi, S., 2013. Development of a coupled
Deckers, J., Haile, M., Nyssen, J., 2012. Impact of conservation agriculture on flow and solute transport modelling for Astaneh-Kouchesfahan groundwater
catchment runoff and soil loss under changing climate conditions in May Zeg-zeg resources, North of Iran. Int. J. Water 7 (1/2), 80. https://doi.org/10.1504/
(Ethiopia). J. Hydrol. 475, 336–349. IJW.2013.051980.
Lawrence, D., Vandecar, K., 2015. Effects of tropical deforestation on climate and Shukla, S., Wood, A.W., 2008. Use of a standardized runoff index for characterizing
agriculture. Nat. Clim. Chang 5 (1), 27–36. hydrologic drought. Geophys. Res. Lett. 35, L02405. https://doi.org/10.1029/
Lima, L. S., Coe, M. T., Soares Filho B. S., Cuadra, S. V., Dias, L. C. P., Costa, M. H., Lima, 2007GL032487.
L. S. , Rodrigues, H. O., 2014. Feedbacks between deforestation, climate, and Skees, J.R., 2000. A role for capital markets in natural disasters: a piece of the food
hydrology in the Southwestern Amazon: implications for the provision of ecosystem security puzzle. Food Pol. 25,365–378.
services, Land. Ecol. 29, 261–274. Soil Conservation Service. 1972. Section 4: Hydrology in National Engineering
Liu, Y., Wang, W., Hu, Y., 2017. Investigating the impact of surface soil moisture Handbook. SCS.
assimilation on state and parameter estimation in SWAT model based on the Stednick, J.D., 1996. Monitoring the effects of timber harvest on annual water yield.
ensemble Kalman filter in upper Huai River basin. J. Hydrol. Hydromech. 65 (2), J. Hydrol. 176, 79–95.
123–133. Swenson, S., Wahr, J., 2009. Monitoring the water balance of Lake Victoria, East Africa,
Liu, Y., Liu, Y., Wang, W., Zhou, H., 2021. Propagation of soil moisture droughts in a from space. J. Hydrol. 370 (1–4), 163–176. https://doi.org/10.1016/j.
hotspot region: Spatial pattern and temporal trajectory. J. Hydrol. 593, 125906. jhydrol.2009.03.008.
Maitima, J.M., Olson, J.M., Mugatha, S.M., Mugisha, S., Mutie, I.T., 2010. Land use Tibebe, D., Bewket, W., 2011. Surface runoff and soil erosion estimation using the SWAT
changes, impacts and options for sustaining productivity and livelihoods in Jthe model in the Keleta Watershed, Ethiopia. Land Degrad. Dev. 22, 551–564.
basin of lake Victoria. J. Sustain. Dev. In Afri. 12, 189–190. Uhlenbrook, B.O., 2004. Quantifying the impact of land-use changes at the event and
Mango, L.M., Melesse, A.M., McClain, M.E., Gann, D., Setegn, S.G., 2011. Land use and seasonal time scale using a process-oriented catchment model. Hydrol. Earth Syst.
climate change impacts on the hydrology of the upperMara River Basin, Kenya: Sci. 8 (1), 62–78.
results of a modeling study to supportbetter resource management. Hydrol. Earth URT, 2001. Tanzania Household Budget Survey 2000/01. Natonal Bureau of Statstcs,
Syst. Sci. 15, 2245–2258. Dar-es-Salaam, Tanzania.
McKay, M.D., Beckman, R.J., Conover, W.J., 1979. Comparison of three methods for van Griensven, A., Ndomba, P., Yalew, S., Kilonzo, F., 2012. Critical review of SWAT
selecting values of input variables in the analysis of output from a computer code. applications in the upper Nile basin countries. Hydrol. Earth Syst. Sci. 16,
Technometrics 21 (2), 239–245. 3371–3381.
McKee, T.B.; Nolan, J.; Kleist, J., 1993. The relationship of drought frequency and Vorosmarty, C.J., Green, P., Salisbury, J., Lammers, R.B., 2020. Global water resources:
duration to time scales. Prepr. Eighth Conf. Appl.Climatol. Amer. Meteor Soc. 17, vulnerability from climate change and population growth. Science 289, 284–288.
179–183. Wagner, P.D., Bhallamudi, S.M., Narasimhan, B., Kumar, S., Fohrer, N., Fiener, P., 2017.
Miller, J.D., Kim, H., Kjeldsen, T.R., Packman, J., Grebby, S., Dearden, R., 2014. Comparing the effects of dynamic versus static representations of land use change in
Assessing the impact of urbanization on storm runoff in a peri-urban catchment hydrologic impact assessments. Environ. Model. Softw. https://doi.org/10.1016/j.
using historical change in impervious cover. J. Hydrol. 515, 59–70. https://doi.org/ envsoft.2017.06.023.
10.1016/j.jhydrol.2014.04.011. Wang, J., Wang, W., Cheng, H., Wang, H., Zhu, Y., 2021. Propagation from
Morris, M.D., 1991. Factorial sampling plans for preliminary computational experiments. Meteorological to Hydrological Drought and Its Influencing Factors in the Huaihe
Technometrics 33 (2), 161–174. River Basin. Water 13(14), 1985; doi: 10.3390/w13141985.
Mutie, S., Mati, B., Gadain, H., and Gathenya, J.: Evaluating land use change effects on Wang, Q., Xu, Y., Gao, B., Wang, Y., Xu, Y., Wu, L., 2017. Runoff spatial responses to land
river flow using USGS geospatial stream flow model in Mara River basin, Kenya, A use change in Xitiaoxi River Basin. In Chinese J. Nat. Resour. 32, 632–641. https://
Paper presentation at the 2nd Workshop of the EARSeL SIG on Land Use and Land doi.org/10.11849/zrzyxb.20160455.
Cover, Bonn, 28–30 September, 2006. Wang, Q., Xu, Y., Xu, Y., Wu, L., Wang, Y., Han, L., 2018. Spatial hydrological responses
Neitsch, S.L., Arnold, J.G., Kiniry, J.R., Williams, J.R., 2011.Soil and Water Assessment to land use and land cover changes in a typical catchment of the Yangtze River Delta
Tool Theoretical Documentation Version 2009. TR-406, Texas Water Resources region. Catena 170, 305–315. https://doi.org/10.1016/j.catena.2018.06.022.
Institute Technical Report No.406. Texax A&M University. (available at http://swat. Woldeamlak, B., Sterk, G., 2005. Dynamics in Land Cover and Its Effect on Stream Flow
tamu.edu/media/99192/swat2009-theory.pdf). in the Chemoga Watershed, Blue Nile Basin, Ethiopia. Hydrol. Process 19, 445–458.
Olivera, F., DeFee, B.B., 2007. Urbanization and its effect on runoff in the White Oak https://doi.org/10.1002/hyp.5542.
Bayou watershed. Texas. J. Am. Water Resour. Assoc. 43, 170–182. https://doi.org/ Woldesenbet, T.A., Elagib, N.A., Ribbe, L., Heinrich, J., 2017. Hydrological responses to
10.1111/j.1752-1688.2007.00014.x. land use/cover changes in the source region of the Upper Blue Nile Basin, Ethiopia.
Rallison, R.E. and N. Miller. 1981. Past, present and future SCS runoff procedure. P. 353- Sci. Total Environ. 575, 724–741.
364. In V.P. Sing (ed.). Rainfall runoff relationship. Water Resources Publication, Yan, R., Zhang, X., Yan, S., Zhang., J., Chen, H., 2017. Spatial patterns of hydrological
Littleton, CO. in upper Huai River basin. responses to land use/cover change in a catchment on the Loess Plateau, China. Ecol.
Remondi, F., Burlando, P., Vollmer, D., 2016. Exploring the hydrological impact of Indic. 92, 151–160. doi: 10.1016/j.ecolind.2017.04.013.
increasing urbanisation on a tropical river catchment of the metropolitan Jakarta, Yevjevich, V., 1967. An Objective Approach to Definitions and Investigations to
Indonesia. Sustain. Cities Soc. 20, 210–221. https://doi.org/10.1016/j. Continental Hydrologic Droughts. Colorado State University, Fort Collins, Colorado.
scs.2015.10.001. Zhang, X., Zhang, L., Zhao, J., Rustomji, P., Hairsine, P., 2008. Responses of streamflow
Rogger,M., Agnoletti, M., Alaoui, A., Bathurst, J.C., Bodner, G., Borga, M., Chaplot, V., to changes in climate and land use/cover in the Loess Plateau China. Water Resour.
Gallart, F., Glatzel, G., Hall, J., Holden, J., Holko, L., Horn, R., Kiss, A., Kohnová, S., Res. 44, W00A07.
Leitinger, G., Lennartz, B., Parajka, J., Perdigão, R., Peth, S., Plavcová, L., Quinton, Zuo, D., Xu, Z., Yao, W., Jin, S., Xiao, P., Ran, D., 2016. Assessing the effects of changes
J.N., Robinson, M., Salinas, J.L., Santoro, A., Szolgay, J.,Tron, S., van den Akker, J.J. in land use and climate on runoff and sediment yields from a watershed in the Loess
H., Viglione, A., Blöschl., G., 2017. Land use change impacts on floods at the Plateau of China. Sci. Total Environ. 544, 238–250. https://doi.org/10.1016/j.
catchment scale: Challenges and opportunities for future research. Water Resour. scitotenv.2015.11.060.
Res. 53, 5209–5219. doi: 10.1002/2017WR020723.

11

View publication stats