ارزیابی تغییرات کاربری زمین با استفاده از داده‌های سنجش از دور چند زمانی و مدل‌های آماری و میزان تاثیر آن در افزایش رواناب (مطالعه موردی حوضه گرگان‌رود استان گلستان)

تغییرات آب و هوا، کاربری زمین و تغییرات پوشش آن، بر فرایندهای رواناب فرسایش تاثیر قابل توجهی دارد. این مطالعه برای  بررسی تغییرات کاربری اراضی و میزان نقش آن در افزایش رواناب در حوضه آبریز گرگان رود طی یک دوره 35 ساله انجام شد و با استفاده از ادغام تکنیک های سنجش از دور و سیستم های اطلاعات جغرافیایی صورت گرفت. در پژوهش حاضر تصاویر لندست مربوط به سال های 1986، 2006 و 2020 پردازش و  تجزیه و تحلیل شد. تاثیر کاربری زمین در پتانسیل رواناب نیز از طریق مدل نیمه توزیعی swat شبیه سازی و برای واسنجی از نرم افزار swat-cup استفاده شد.. نتایج حاصل از بررسی های انجام شده در تغییرات کاربری اراضی طی بازه زمانی مشخص، حاکی از آن است که در بازه زمانی 2020-1986، بیشترین تغییر شکل مربوط به اراضی جنگلی است که از3342.80 کیلومتر مربع در سال 1986 به 1819.87کیلومتر مربع در سال 2020 کاهش یافته است. نتایج به دست آمده نشان می دهد که تغییرات کاربری اراضی در حوضه رودخانه گرگان رود، به افزایش حجم رواناب از 149.45در سال 1986 به 159.86 میلی متر در سال 2020 (به میزان 6.89درصد افزایش داشته ) منجر شده است و بیشترین تاثیر در افزایش رواناب نیز مربوط به افزایش کاربری اراضی کشاورزی و کاهش اراضی جنگلی به ترتیب به میزان 43.63 و 37.25درصد، در طی دوره آماری 2020-1986است. در نهایت، مقایسه اثر تغییرات کاربری اراضی بر مولفه های چرخه هیدرولوژیکی و مقادیر مختلف رواناب نشان می دهد که با ادامه روند تخریبی کاربری، مقادیر مختلف رواناب افزایش یافته و نفوذپذیری و آب گذری به آبخوان های سطحی و عمیق نیز کاهش یافته است؛ این امر بر توان سیلاب در حوضه مورد مطالعه تاثیر می گذارد. نتایج این مطالعه، برای مدیریت حوادث شدید و برنامه‌ریزی/ مدیریت کاربری اراضی در آینده در منطقه مرتبط و مفید است

the relationship between land use changes in the increase of runoff in the gorgan river watershed using satellite image data and statistical models

1- introduction  land use and land cover (lulc) is a complex set of changes caused by the interaction of the natural environment and human activities, which has an important impact on the global environmental changes and sustainable development (li et al., 2020). changes in land use can occur due to population growth and the development of regional activities (prayitno et al., 2020). many of the problems caused by this development can be soil erosion, soil degradation, and the reduction of forest areas and biodiversity (hu et al., 2019), which have had a major impact on the regional and global environment (chen et al., 2020). changes in lulc can not only directly affect the quantity and quality of land resources in human life, but also indirectly cause climate change, which is one of the important factors of global warming (baoying et al., 2008), so it can change the hydrological regime and rainfall-runoff mechanisms of a region (li et al., 2007). factors such as land use changes, rainfall intensity, and degree of soil saturation, etc. cause the balance and natural flow of rivers to be disrupted (ghanavati et al., 2014). the expansion of urbanization leads to an increase in impervious areas and a decrease in rainfall absorption in the watershed, causing changes in the river’s hydrology, creating runoff after rainfall, and as a result, reducing the recharge of the aquifer (quan et al., 2015). around the world, flood includes almost one-third of natural hazards and harms people more than any other types of disasters (asinya et al., 2021). change detection of lulc is possible by comparing the changes that occurred in a certain area according to the images taken at different times. today, satellite data on land resources are available and are relevant and useful for lulc studies (shanmugapriya et al., 2016) because of having some features such as high temporal frequency, accessibility, showing global land cover for consecutive years, being suitable for calculations, and having a wide range of uses which make them have a high potential for analyzing spatial and temporal changes (kantakumar, 2019). in recent years, due to climatic reasons, the occurrence of floods has increased in the world (ghanavati et al, 2013). golestan province is no exception to this rule. in the recent floods in golestan, natural factors such as the wet winter have led to the wetting of the soil, the filling of storage channels, and the rise of the stagnation level, and as a result, the runoff coefficient has increased. in terms of human factors, it is possible to point out the impact of non-observance of the principles of land preparation and improper land use allocation, deforestation, encroachment on river boundaries, and insufficient dredging of the main channels especially the estuary, which have increased the possibility of occurrence of natural hazards. hydrological models are the basis for understanding the cause-and-effect relationship between hydrological changes and land use changes (shokouifar et al., 2022).2- methodology  in this research, landsat tm, landsat etm+, and tirs oli satellite images from 1986, 2006, and 2020 have been respectively used to classify and investigate land use changes in the gorgan river basin. envi 5.6 software was used for image processing and data analysis, and arcgis 10.7 software was used to obtain output from image processing. envi 5.6 software was used to classify the desired images using the random forest algorithm and using the enmap-box 2.2 plugin. terrset2020 software was used to model the changes. the kappa coefficient was used to evaluate the accuracy and precision of the classification as well as to compare the classification result with the ground reality. for land cover classification, six land use classes, including forest, urban areas, agricultural lands, water areas, pastures with good vegetation cover, and land with poor cover (pasture and barren land) were considered. in this study, the effect of land use on runoff potential was also simulated through the semi-distributed swat model. model implementation was done in arc gis 10.7 environment. after preparing the required maps and preparing the input data, three different swat models were designed for the gorgan river watershed. the first model was used from 1985 to 1996 from the land use map of 1986, the second model was used from 1999 to 2009 from the land use map of 2006, and the third model was used from 2010 to 2020 from the land use map of 2020. in the first stage, by entering the dem map and producing the flow network by the model itself, based on the threshold limit of 14,000 hectares as the minimum drainage level and entering the agh qala hydrometric station as the outlet of the basin, the gorgan river watershed was divided into 32 sub-basins. after drawing the boundary of the basin, sub-basin, and flow network, the physical parameters related to the basin and each sub-basin, including area, length of the main waterway, slope, height characteristics, etc. are calculated. in the next step, the soil and land use maps must be entered into the model and the slope classes must be defined by the user, and by combining them, hydrological reaction units (hru) are produced in each sub-basin. the number of hrus can be changed multiple times for each land use, soil, and slope by determining a minimum percentage of the watershed area that is defined by the user. in the next step, climate data including daily precipitation and temperature information are entered into the models and the appropriate method for calculating potential evaporation and transpiration is determined based on the type of climate data available. in this study, the hargreaves-samani method was used to calculate potential evaporation and transpiration. the method of variable storage coefficient was used for trending the flow. also, management information such as planting, fertilizing, irrigation time, and harvesting of the dominant crops of the basin were introduced to the models. in the final step, the model was run to simulate monthly runoff, considering 3 years of training for all three models.3-results the analysis of annual runoff in three scenarios shows that under the second and third scenarios, the surface runoff has increased by 20.47 and 46.45%, respectively, compared to the first scenario according to the investigations, it is clear that water efficiency has been increasing from 1986 to 2020. this increase can be attributed to land use changes, including the reduction of forest area and the increase of agricultural land, pastures, and residential areas. an increasing trend is observed in the northeast sub-basins compared to the southwest, which is due to the reduction of forest land and its conversion to agricultural land in the northeast. in 1986, the water yield of most sub-basins (45.98% of the basin area) is less than 194 mm. while the water yield in 2020 has increased by more than 290 mm in most of the basins (56% of the basin area) and in sub-basins 10 and 12, which are mainly degraded forest areas and are at a higher level up to 378mm, the variable of agricultural land with an average participation rate of 43.63% has had the highest change in runoff from 1986 to 2020. and after that, forest lands increased by 37.25% and played an important role in creating runoff in the gorgan basin.4- discussion & conclusionthe results show that urban land has increased from 3.20% of the total land of the region in 1986 to 4.66% in 2006 and this number has reached 5.60% in 2020. according to the investigations carried out in this