Changes Of Hydrological Elements And Effects Of The Naoli River Watershed In The Sanjiang Plain, Northeast China

Sanjiang Plain, locating in Heilongjiang province of Northeast China, coveredthe largest area of fresh water marshs of China. With construction of manystate-owned farms, Sanjiang Plain has experienced great agricultural developmentand extensive agricultural land expansion since the1950s. Widespread wetlands inand around the farms has experienced shrinkage and fragmentation. The unbalancebetween ecological water requirements of wetland ecosysterm and agricultural waterrequirements was significant. The Naoli River, locating in the center of the SanjiangPlian, was typical marsh river. However, the hydrological elements of Naoli Riverwatershed has changed significantly because of the firece variation of lands usepattern. The objects of the paper were to analysize the change of hydrologicalelements and its eco-environmental effect induced by the variation of lands usepattern in the Naoli River watershed. The methods used in the paper are the SWAT(Soil and Water Assessment Tool) model and the Range of Variability Approach(RVA) based on the land use maps and hydo-meteorological data and the majorconclusions are as follows:(1) Analysizing the variation of hydo-meteorological elements during1956~2008using mathematical statistical methods in the Naoli River watershed. Theaverage annual runoff was decreased notably and the abrupt change of runoffoccurred in1966. The distribution of seasonal runoff was quite uneven. The floodevents in summer were larger than that in spring. The flood frequency curves weredifferent between the Baoqing station and the Caizuizi station. Moreover, the average annual precipitation and ET all decreased while the temperature increased inthe past sixty years. However, the trends of decreasing or increasing wereunsignificant which suggested that the meteorological elements were not the mainfactors leading to the runoff reduction.(2) Large scale land reclamation activites have happened and wetlands havebeen shrinking following the farmland expanding since the1950s. The area ofwetlands was9453.4km2in1954and decreaed to1837.2km2in2005. Thepercentage of wetland area occupied with total area of the Naoli River watershedreduced by32.2%. In contrast, the area of farmland was3287.7km2in1954andincreased to14699.1km2in2005. The percentage of the farmland occupied withtotal area of the Naoli River watershed increased by48.2%. Also, the water storagefunctions of wetlands were calculated based on field sampling and lands use maps.In1954and2005, the soil water-holding capacities were52.7×108m3and10.2×108m3within the100cm soil layer, respectively. Since marshes distribute atlow land, average water level is15cm, then the water storage amount were18.4×108m3and3.7×108m3in1954and2005, respectively. Therefore, the total waterstorage capacities of the wetlands were71.1×108m3and13.9×108m3in1954and2005, respectively.(3) Ascertain the sensitivity parameters of the SWAT model according to thedifferent characteristics of the terrain and natural features in the upper and middle ofthe Naoli River watershed. The sensitivity parameters in Caizuizi station includedCN, SOL_AWC, ESCO, and CH_K2, et al, which were the same as that in Baoqingstation. In addition, the SURLAG, GWQMN, REVAPMN and wetland parameterswere sensitivity parameters as well in Caizuizi station. Because of no or sparsehydro-meteorological data, the calibrated SWAT model was applied to the QixingRiver watershed according to hydrological similarity and model parameterstransplantation method. The similarity degree between the upper Naoli Riverwatershed and the Qixing River watershed was0.910. By the model effectivenesstesting, the values of NS and R2were both larger than0.64, and PBIAS was smallerthan9for calibration and validation periods. Above all, the SWAT model was appropriate to simulate the runoff of the Naoli River watershed.(4) Making land use change scenarios based on lands use pattern change inorder to analysize the effects of two lands use transformations, that were fromwetland to farmland and from forest to farmland to runoff and ET in the Naoli Riverwatershed. When wetland area decreased by38%, the average annual runoff reducedby36.73%while the average annual ET increased by3.39%. When wetlandtransfromed into farmland totally, the average annual runoff decreased by49.18%corrisponding with the increasing rate of16.78%of ET. Besides, the effects of landuse change between forest and farmland was analysed in the upper Naoli Riverwatershed. The forest decreased by90.81%along with the runoff raisng of83.48%and ET reduction of0.04%.(5) The annual runoff series in the Naoli River watershed were divided into twostages namely the period of more natural or less altered flow conditions (1956-1966)and the period of more altered conditions (1967-2005) according to theMann-Kendall test. Computing the33IHA (Indicators of Hydrologic Alteration)factors in both Baoqing station and Caizuizi station with the RVA (Range ofVariability Approach), and the degrees of comprehensive hydrologic alteration were0.69and0.68, respectivily, indicating high alteration of hydrological regime.