Evaluation of the Hydrological Response of Kiwi Chai Catchment Area to Land Use Changes Using SWAT Model

Land use change has a direct impact on hydrological components and water resources and plays an important role in aggravating possible risks such as drought and floods. Therefore, it is necessary to investigate the effects of land use on water components such as runoff. Thus, in this study, the runoff condition of the Kiwi Chai Basin, one of the most important basins in Ardabil province, from an environmental point of view in terms of land use change is investigated using the SWAT model.

ArcGIS 10.1, Envi 5.1, and Ecognition software are used for data processing, classification, and analysis. Also, the method based on radiative transfer models, such as FLAASH, is used as the best method for atmospheric correction. Multiresolution algorithm is applied for segmentation and the nearest neighbor algorithm is used for classification. Then, educational samples are collected from the area using field visits and Google Earth satellite images and used to evaluate the accuracy and precision of the classification. At the end, the Thematic change dynamic detection method is applied in Envi5 to identify the changes that have happened over the years. The SWAT hydrological model is additionally used to investigate the effect of land use changes on the basin's runoff. After extracting the effective parameters in the basin’s outlet, the model is calibrated and validated with SWAT-CUP software and SUFI-2 algorithm. Nash-Sutcliffe indices and coefficient of determination are used to evaluate the results of this stage.

In this research, after applying object-oriented land classification, land use map in seven classes of Rainfed agriculture, irrigated agriculture, garden, forest, residential, pasture and water are provided. In the land use map of the basin, pasture cover is the dominant land use class; however, a significant decrease from 1224.18 square kilometers to 1046.59 square kilometers has been observed between 1987 and 2019. The values obtained for R-Factor, P-Factor, R2, NS indicators in the calibration period are, respectively, for Abgarm station (0.53, 0.47,0.71,0.69) and for Firozabad station (0.32,0.3,0.67,0.64) and in the validation period, for Abgram station (0.09,0,0.62,0.56) and for Firozabad station (0.13,0,0.53,0.51), respectively. Based on this, the results obtained in the calibration and validation stage are evaluated as acceptable. The evaluation of the SWAT model's response to land use also shows that, on an annual scale, the amount of flow in Firozabad station has decreases from 3.08 to 2.81 cubic meters per second -8,77percent) and in Abgarm station from 1.11 to 0.96 cubic meters per second (-13.51), which can be attributed to changes in land use, especially the change in the use of pastures and its conversion to rainfed agriculture, gardens, and forests.

The trend of land use changes in the basin between 1987 and 2019 has been accompanied by a decreasing trend in rainfed, pasture and water uses and an increasing trend in irrigated agriculture, garden, forest, and residential uses. The high value of the fit of the indicators used in the evaluation of the model indicates that the model has a good capability in simulating the runoff of the basin. Moreover, the model implemented for different land uses illustrates that the flow of the basin in both hydrometric stations in most of the months increase in the short term and decrease in the long term average with the land use change. Continuous land use change is becoming a serious threat to watersheds. Land use change should be controlled in catchment areas, and measures should be taken to stabilize land use change.