نوع مقاله : پژوهشی
نویسندگان
1 گروه مهندسی آبخیزداری، دانشکده منابع طبیعی، دانشگاه علوم کشاورزی و منابع طبیعی ساری، ساری، ایران
2 استاد گروه آبخیزداری، دانشگاه علوم کشاورزی و منابع طبیعی ساری
3 گروه مهندسی آبخیزداری، دانشکدۀ منابع طبیعی ساری، دانشگاه علوم کشاورزی و منابع طبیعی ساری، ایران
4 استادیار گروه علوم و مهندسی آبخیزداری، دانشکده منابع طبیعی و علوم دریایی دانشگاه تربیت مدرس. تهران، ایران.
چکیده
کلیدواژه‌ها
موضوعات
عنوان مقاله [English]
نویسندگان [English]
Introduction and purpose
Studies of land use and land cover plays an important role in planning for resource management and protection and provides the basis for improving the fundamental attitude to environmental structures. Urban and agricultural development has a great impact on land use and land cover. For this reason, in recent years, more than half of the earth's surface has undergone changes, and more than one-third of the earth's area has been devoted to agricultural lands. As a result of these major changes, land use managers and experts have investigated the impact of land use change on hydrological processes. Machine learning techniques, such as artificial neural network (ANN), support vector machine (SVM), random forest (RF), decision tree (DT) and other models, have attracted much attention for land cover/use (LULC) classification. Planners and managers can use predicted LULC changes to promote sustainable land management and reduce adverse impacts. As a result, the detection and prediction of land use changes (LULC) caused by rapid urbanization can lead to disruption of environmental sustainability. Population growth, economic development, and agricultural expansion are factors that cause various changes in land covers, including vegetation and water. These continuous changes in land use can lead to environmental degradation. The intensity of these changes in response to the growth of the world's population and the increase in the need for food shows the necessity of conducting detailed studies about these changes.
The purpose of this study is to observe land use changes in 2000, 2014 and 2024 in Darab region. This research also focused on identifying the driving force for LULC changes. The cell-automated-artificial neural network (CA-ANN) model was also conducted to analyze the predicted patterns and trends of land use from 2034 to 2044.
Materials and methods
Land use classification of satellite images was done by implementing pixel-based classification technique and monitoring in Google Earth Engine environment. Land use classification was done through SVM support vector machine algorithm. After the analysis of the studied area, there were five different land use classes including pastures, barren lands, gardens, agriculture and urban areas. It is suitable for assessing spatial and temporal land use changes, modeling transfer potential and predicting future scenarios. Artificial Neural Network (ANN) in conjunction with Cell Automatization (CA) was used to predict land use changes. The MOLUSCE plugin was implemented in QGIS to generate spatio-temporal changes with a time period of 2034 to 2044 and calculate LULC transitions to create a LULC change map. Also, a transition potential matrix between 2000-2014 was created to create a change map. The multi-layer perceptual neuron (ANN-MLP)-ANN method was used to model the transmission potential. Slope, direction, height and distance from the road, fault and river were the spatial parameters that were implemented as input parameters. ANN-MLP architecture, where the input layer is processed by hidden layers and the output layer contains the reclassified LULC classes.
Results and discussion
Land use change (LULC) maps prepared from 2000 to 2024 show a significant increase in the area of agricultural land in the region. The natural ecosystems of this plain are facing serious challenges due to the growing trend of land use changes and exchanges. These changes are caused by unprincipled and irrational exploitation of resources as a result of human activities such as urbanization, improper agriculture, digging wells and excessive use of underground water resources for cultivating gardens. These factors can lead to erosion and desertification in this region and are considered a serious threat to the environment.
Due to the change in the use of agricultural land, it is faced with a large expansion along with significant changes in barren land. Also, the artificial neural network-automated cell (CA-ANN) technique was used to predict LULC changes for the period 2044-2034. The simulation accuracy percentage was 82.43% and the overall kappa value was estimated at 0.72. The forecast maps from 2034-2044 showed continuous growth in the agricultural land use pattern. In this case, the percentage of area changes increased from 2034 with 455.65 square kilometers (18.52%) to 708.81 square kilometers (28.81%) in 2044. According to the observed results, physical and socio-economic factors had a significant effect on landscape patterns during the study period. The geographic variables included in the model calibration were selected because of their significant relationship with LULC. Physical variables, such as geography and climate, are thought to be the most important factors in promoting human activities. Socio-economic factors, such as population and GDP, may influence LULC change.
Conclusion and suggestions
Changing LULC patterns can negatively affect groundwater quality and also endanger food security. For LULC classification, Support Vector Machine (SVM) provided detailed representations of land cover changes and related trends. In the end, according to the results and findings obtained from the present research, it should be mentioned that any change in land use should be based on scientific principles, with a plan and according to logical justifications; So that it is necessary to use up-to-date and efficient methods such as remote sensing techniques and geographic information systems; Therefore, it is hoped that the future prediction map of the land use situation prepared as a guide for land use planning will be considered and used by the relevant planners, officials and operators, so as to avoid irreparable environmental damage (Employment of underground water sources, desertification, land subsidence). Darab plain should be prevented in the future. Due to the increase in the rate of the area covered by agricultural use, a decreasing trend was observed for barren lands in the simulated maps. The driving factor of land use change in Darab depends on the rapid rate of population growth and increase in demand and conversion of other uses to agriculture, gardens and residential areas.
کلیدواژه‌ها [English]
)