بررسی اثر تغییرات کاربری اراضی روی رواناب حوضۀ رودخانۀ قره چای با استفاده از مدل swat

در پژوهش حاضر، تغییرات کاربری ارضی و تاثیر آن روی رواناب حوضۀ قره‌چای در استان همدان با استفاده از مدل swat تجزیه‌وتحلیل شده است. در این پژوهش، به‌منظور بررسی تاثیر کاربری اراضی بر رواناب حوضۀ آبخیز قره‌چای، از دو تصویر ماهواره‌ای لندست oli-tm سال‌های (2001 تا 2020) استفاده شد. ابتدا تصاویر مربوطه اخذ و پیش‌پردازش‌های لازم شامل تصحیحات اتمسفری به روش flaash  اعمال شد. برای افزایش دقت طبقه‌بندی روش تلفیق تصاویر چند طیفی با تصویر پانکروماتیک[1] انجام شد و قدرت تفکیک مکانی به 15 متر ارتقا یافت. سپس طبقه‌بندی با استفاده از روش شیءگرا و الگوریتم نزدیک‌ترین همسایگی صورت گرفت. از مدل swat برای شبیه‌سازی هیدرولوژیکی حوضه استفاده و از الگوریتم sufi-2 در نرم‌افزار swat-cup برای تحلیل حساسیت، واسنجی و اعتبارسنجی بهره گرفته شد. با توجه به حساسیت مدل به پارامتر تلفات اولیه، واسنجی مدل براساس مقادیر تلفات اولیه انجام شد. مقادیر این ضرایب برای دورۀ واسنجی بین 0.72 تا 0.90 و نتایج اعتبارسنجی مدل، نشان‌دهندۀ تایید صحت واسنجی انجام‌شده بود. بررسی کارایی مدل با استفاده از ضرایب ناش ساتکلیف، p-factor،  rو r-factor حاکی از قابلیت زیاد مدل در شبیه‌سازی رواناب است.

investigating the effects of land use changes on the runoff of qara chai river basin using the swat model

abstract:the present study analyzed changes in land use and its effects on the runoff of qara chai watershed in hamadan province using the swat model. to this aim, landsat oli-tm satellite images of the years of 2001-2020 were used. first, the relevant images were obtained and the necessary pre-processing steps, including atmospheric corrections, were applied by using the flaash method. to increase the classification accuracy, the multispectral images were combined with the panchromatic images and the spatial resolution was enhanced up to 15 m. then, the classification process was done by using the object-oriented method and the nearest neighbor algorithm. the swat model was utilized for hydrological simulation of the basin and sufi-2 algorithm was applied in swat-cup software for conducting sensitivity analysis, calibration, and validation. due to the sensitivity of the model to the initial loss parameter, it was recalibrated based on the initial loss values. the values ​​of the coefficients for the calibration period were between 0.72 and 0.90 and the model validation results confirmed the calibration accuracy. assessment of the efficiency of the model by using nash sutcliffe coefficients, p-factor, r, and r-factor indicated its high capability for simulating the related runoff.keywords: land use, runoff, swat modelintroductionland use and its fluctuations are among the factors that affect natural cycles in ecosystems. land use change has several effects on the hydrology of watersheds, such as changing the characteristics of peak discharge, total volume of runoff water, water quality, and hydrological balance. changing land use and soil cover has some effects on runoff, infiltration, retention, evaporation, transpiration, etc. the runoff caused by rainfall in watersheds causes erosion and loss of surface fertile soil and finally leads to sedimentation in canals, rivers, and reservoirs of dams in addition to human and financial losses triggered by floods. nowadays, it is possible to study land use changes through satellite images and remote sensing science the geographic information system as a suitable tool for extracting and classifying information has greatly contributed to the studies related to the effects of land use changes.swat model is a hydrological simulator and a continuous spatio-temporal semi-distributed model with a physical base. it has been used to simulate hydrological processes in complex and vast watersheds with regard to soil changes, land use, and weather conditions during different periods for a long time. methodologyqara chai basin with an area of ​​about 11000 km2 (based on the political border) is the largest watershed in hamadan province. it has got the geographical coordinates of 13° 48° to 29° 49° east longitude and 12° 34° to 35° 44° north latitude. it is drained by qara chai river and its branches. this basin is the most flood-prone part of hamadan province in terms of flood situation.in the studied plan, the area of qara chai catchment (​​11000 km2) was divided into 18 sub-basins. the swat model was used to obtain the data of the land use, soil, and slope maps and build the hydrological response units. separate grids were introduced to the model. after introducing two maps of land use, as well as soil and slope classes, to the swat model, all the rasterized maps were combined and integrated to produce the hydrological reaction units of the basin. due to the fact that the land use maps had changed in two time periods and thus had different characteristics, the hydrological response units were affected by the land use type in addition to the soil type and slope. therefore, therefore, the number of the hydrological reaction units changedafter forming the basin and hydrological response units, the meteorological data required by the swat model were introduced to the model. then, the output of this model was linked with sufi-2 program in swat-cup software and the important parameters of the swat model for the watershed were sensitized with the help of sufi-2 algorithm for the statistical period of 2002-2021 by applying the method of one parameter at a time. the parameters, which were more sensitive, were identified and their optimal values ​​were determined. in short, the model was recalibrated for the land uses of 2001 and 2020 during the statistical periods of 2001-2008 and of 2013-2020, respectively. it was validated for the statistical period of 2009-2012. using the statistical criteria of nash sutcliffe coefficient, p-factor, and r-factor both for the calibration and validation periods, the statistical analysis was done and the statistical results were checked by using the swat-cup program. results and discussionin the present research, land use changes and its effects on the runoff of qara chai basin were analyzed by using the swat model. to simulate the runoff by using the mentioned model and sufi-2 algorithm, the effective parameters on the runoff were identified from among the 29 parameters examined in the sensitivity analysis via the sensitivity analysis technique. 14 variables were found to affect the simulation.