Modelling Evapotranspiration Processes At Different Scales In Qinhuai River Basin

As a result of the expansion of urbanization by human activities,amount of crop land convert to urban use and then further affect the eco-hydrological cycle together with climate change.Evapotranspiration(ET)is a key to connect the water,energy and carbon cycle.It represents the balance of material and energy information.Therefore,studying how climate change and land-use/land-cover change(LUCC)affect evapotranspiration by integrating multiple techniques of remote sensing,hydrological modeling,and ground-based observations is essential and beneficial for explaining hydrological processes in a rapidly urbanizing subtropical humid region.A typical rapidly urbanized watershed in southern China,Qinhuai River Basin(QRB),is selected as the study area.This study analyzed the distribution of different land-cover types of QRB based on land-use/land-cover map of 2000,2004 and 2011 derived from remote sensing.We revised the Soil and Water Assessment Tool(SWAT)by incorporating new algorithms for modeling hydrological processes of rice paddies.Observed streamflow during 1990-1999 of Wuding sluice gate and ET measurements in 2007 at a rice paddy field with the weighing lysimeter method was used to for model calibration and validation.Simulation and discussion of ET variation in different time scales(yearly,monthly and seasonal),land-cover types and spatial scales(watershed scale and HRU scale)have been conducted from 2000-2013.Furthermore,we separated the effects of LUCC from climate at the watershed scale.(1)The QRB has experienced high rate of urbanization from 2000-2013 where urban use area increased by 11.3%,rice paddy field,agricultural land,forest and water decreased by 7.6%,2.4%,0.5%and 0.6%,respectively,resulting in converting crop land especially paddy rice field to urban use.(2)The SWAT has been improved by modifying the algorithm of surface area,irrigation and evaporation of pothole module.The original SWAT model showed negative correlations between the observed and simulated streamflow.NSE were lower than 0 both in 1990-1994(NSE =-3.28)and 1995-1999(NSE =-8.42),and the simulated value were much higher than the observations.However,the results based on the performance indicators showed a good agreement between measured and modeled streamflow by the improved SWAT model after calibration and validation.The Nash-Sutcliffe model efficiency(NSE)was 0.86 and R2 was 0.88 for the calibration period(1990-1994)while the NSE was 0.65 and R2 was 0.71 for the validation period(1995-1999).We also found good agreements between modeled daily ET by improved SWAT and lysimeter measurements for the rice paddy filed(R2 = 0.75 and p<0.01),while the original SWAT showed low relationship for R2 = 0.49.It demonstrates that the improved SWAT model is better for applicability.(3)Overall,simulated annual actual ET at the watershed scale decreased from 2000-2013 obviously,which is contrary to increased PET and VPD.Variations of ET,PET and VPD follow similar seasonal patterns at monthly scale.The values increased since January and reached to the maximum in July/August,and then decreased gradually.At seasonal scale,ET rates decreased in all seasons mostly due to reduction of rice paddy field in growing season.This proved that LUCC of converting rice paddy field to urban use is the main factor to the total ET decrease.(4)The driving factors to ET change vary with spatial scale.For watershed scale,ET of rice paddy field accounts for the largest proportion of the total ET,while the proportion of urban use is the smallest.The rice paddy,agricultural and forest ET is decreasing(6.8%,1%and 1.2%),while ET of urban use increased by 9.1%,illuminating that LUCC plays the principal role at watershed scale.At HRU scale,ET of all land-cover types is increasing significantly,especially for forest and rice paddy field.It demonstrates that ET variation is dominated by meteorological factors at HRU scale.(5)The deviation of simulated ET indicated that LUCC contributed to the decrease of ET by about 53%to 61%of the observed decrease in ET.This was in contrast to the role of climate variability that contributed to ET increased from 47%to 39%.Overall,the combined accumulated contributions of climate and LUCC caused the annual ET to decrease by 21%.It illuminates that ET reduction is not only influenced by climate change,but also strongly affected by LUCC,and LUCC is the main driver to ET variations.