نشریه علوم زمین خوارزمی

نشریه علوم زمین خوارزمی

تحلیل همبستگی پوشش گیاهی و رطوبت سطحی با استفاده از داده‌های سنجش از دور و مدل‌های یادگیری عمیق (مطالعه موردی: حوزه دریاچه ارومیه)

نویسندگان
دانشگاه شهید بهشتی
چکیده
پوشش گیاهی یکی از عناصر کلیدی در حفظ تعادل اکولوژیکی و پایداری منابع طبیعی است که نقش مؤثری در تنظیم چرخه آب، کاهش فرسایش خاک و حفظ تنوع زیستی ایفا می‌کند. این پوشش به‌طور مستقیم با رطوبت سطحی خاک در تعامل است و تغییرات آن می‌تواند نشانه‌ای از دگرگونی‌های اکولوژیکی در محیط‌های حساس مانند دریاچه ارومیه باشد. بررسی همبستگی بین پوشش گیاهی و رطوبت سطحی برای درک بهتر روندهای محیط‌زیستی و مدیریت پایدار منابع آبی ضروری است. در این پژوهش، با بهره‌گیری از روش‌های یادگیری عمیق و تحلیل شاخص‌های سنجش از دور NDVI و NDWI تغییرات پوشش گیاهی و رطوبت سطحی در چهار ناحیه اطراف دریاچه ارومیه طی بازه‌های ۲۰۱۵ تا ۲۰۲۴ بررسی شده و مقادیر سال۲۰۲۵ پیش‌بینی گردید. مدل‌سازی با استفاده از شبکه‌های عصبی عمیق انجام شده و دقت آن با معیار ریشه میانگین مربعات باقی‌مانده‌ها (RMSE) ارزیابی شد. تحلیل آماری نتایج نشان داد که نواحی جنوبی و غربی بیشترین نوسانات را تجربه کرده‌اند. همچنین، ضریب همبستگی بین NDVI و NDWI در تمامی نواحی منفی و معنادار بوده (r بین 0.83- تا 0.96-، p < 0.01) که نشان‌دهنده تأثیر منفی افزایش پوشش گیاهی بر رطوبت سطحی است. این یافته‌ها ضمن تأیید وجود یک رابطه معکوس قوی، بیانگر اهمیت تلفیق داده‌های پوشش گیاهی و رطوبت در پایش بلندمدت، پیش‌بینی تغییرات اکولوژیکی و اتخاذ راهکارهای مدیریتی مؤثر برای احیای دریاچه ارومیه می‌باشند.
کلیدواژه‌ها

عنوان مقاله English

Correlation analysis of vegetation cover and surface moisture using remote sensing data and deep learning models (case study: Lake Urmia area)

نویسندگان English

Amir Ghodsifar
Asghar Milan
Shahid Beheshti University
چکیده English

Vegetation cover is one of the key elements in maintaining ecological balance and the sustainability of natural resources, playing an effective role in regulating the water cycle, reducing soil erosion, and preserving biodiversity. This cover interacts directly with surface soil moisture, and its changes can indicate ecological transformations in sensitive environments such as Lake Urmia. Examining the correlation between vegetation cover and surface moisture is essential for a better understanding of environmental trends and the sustainable management of water resources. In this study, using deep learning methods and the analysis of remote sensing indices NDVI and NDWI, changes in vegetation cover and surface moisture in four regions around Lake Urmia from 2015 to 2024 were investigated, and the 2025 values were predicted. Modeling was conducted using deep neural networks, and its accuracy was evaluated using the Root Mean Square Error (RMSE) criterion. Statistical analysis of the results showed that the southern and western regions experienced the most fluctuations. Furthermore, the correlation coefficient between NDVI and NDWI across all regions was negative and significant (r between -0.83 and -0.96, p < 0.01), indicating the negative impact of increasing vegetation cover on surface moisture. These findings, while confirming a strong inverse relationship, highlight the importance of integrating vegetation cover and moisture data in long-term monitoring, predicting ecological changes, and adopting effective management strategies for the restoration of Lake Urmia.

کلیدواژه‌ها English

Vegetation Cover
Surface Moisture
Correlation
Deep Learning
Remote Sensing
Aa, Y., Wang, G., Hu, P., Lai, X., 2022. Root-zone soil moisture estimation based on remote sensing data and deep learning. Environmental Research 212, 113278.
Abbes, A., Noureddine, J., Farah, I., 2024. Advances in remote sensing-based soil moisture retrieval: applications, techniques, scales and challenges for combining machine learning and physical models. Artificial Intelligence Review 57, 1–22.
Amgoth, A., Rani, H., Kv, J., 2021. Monitoring of Dynamic Wetland Changes using NDVI and NDWI based Landsat Imagery. Remote Sensing Applications: Society and Environment 23, 100547.
Azizi, S., 2024. Drought and Environmental Transformations of Lake Urmia: An Integrated Analysis Using Machine Learning and GIS. Advances in Image and Video Processing 12, 290–315.
Babazadeh, H., Norouzi Aghdam, A., Aghighi, H., Shamsnia, S.A., Khodadadi Dehkordi, D., 2012. Estimation of surface soil moisture of rangelands in arid and semi-arid regions using the temperature–vegetation index (case study: Khorasan Province). Iranian Journal of Range and Desert Research 19, 120–132. (in Persian)
Bandak, S., Naeini, S., Komaki, C.B., Verrelst, J., Kakooei, M., Mahmoodi, M., 2023. Satellite-Based Estimation of Soil Moisture Content in Croplands: A Case Study in Golestan Province, North of Iran. Remote Sensing 15, 2155.
Cai, Y., Liu, S., Lin, H., 2020. Monitoring the Vegetation Dynamics in the Dongting Lake Wetland from 2000 to 2019 Using the BEAST Algorithm Based on Dense Landsat Time Series. Applied Sciences 10, 4209.
Çelik, M., Isik, M.S., Yuzugullu, O., Fajraoui, N., Erten, E., 2022. Soil Moisture Prediction from Remote Sensing Images Coupled with Climate, Soil Texture and Topography via Deep Learning. Remote Sensing 14, 5584.
Emami, H., Zarei, A., 2021. Modelling lake water's surface changes using environmental and remote sensing data: A case study of lake urmia. Remote Sensing Applications: Society and Environment 23, 100594.
Feyisa, G.L., Meilby, H., Fensholt, R., Proud, S., 2014. Automated Water Extraction Index: A New Technique for Surface Water Mapping Using Landsat Imagery. Remote Sensing of Environment 140, 23–35.
H.S. Asadollah, B., Sharafati, A., Saeedi, M., Shahid, S., 2023. Estimation of soil moisture from remote sensing products using an ensemble machine learning model: a case study of Lake Urmia Basin, Iran. Earth Science Informatics 17.
Herati, J., Kiadaliri, M., Tavana, A., Rahnavard, A., Amirnejad, R., 2023. Relationship between changes in water extent and vegetation cover in eastern Lake Urmia and the dust storm phenomenon. Journal of Civil and Environmental Engineering 53, 44–54. (in Persian)
Kazempour Choursi, S., Erfanian, M., Ebadi Nahari, Z., 2019. Evaluation of MODIS and TRMM satellite data in drought monitoring of the Lake Urmia watershed. Environmental Geography and Planning 30, 17–34. (in Persian)
Kazemi Garajeh, M., Malakyar, F., Weng, Q., Feizizadeh, B., Blaschke, T., Lake, T., 2021. An automated deep learning convolutional neural network algorithm applied for soil salinity distribution mapping in Lake Urmia, Iran. Science of The Total Environment 778.
Khoshnood, S., Lotfata, A., Sharifi, A., 2022. Unsustainable Anthropogenic Activities: A Paired Watershed Approach of Lake Urmia (Iran) and Lake Van (Turkey). Remote Sensing 14, 5269.
Li, M., Yan, Y., 2024. Comparative Analysis of Machine-Learning Models for Soil Moisture Estimation Using High-Resolution Remote-Sensing Data. Land 13, 1331.
Li, S., Xu, L., Jing, Y., Yin, H., Li, X., Guan, X., 2021. High-quality vegetation index product generation: A review of NDVI time series reconstruction techniques. International Journal of Applied Earth Observation and Geoinformation 105, 102640.
Meng, X., Zeng, J., Yang, Y., Zhao, W., Ma, H., Letu, H., Zhu, Q., Liu, Y., Wang, P., Peng, J., 2024. High-resolution soil moisture mapping through passive microwave remote sensing downscaling. The Innovation Geoscience 2, 100105.
Mirzaee, S., Nafchi, A.M., Ostovari, Y., Seifi, M., Ghorbani-Dashtaki, S., Khodaverdiloo, H., Chakherlou, S., Taghizadeh-Mehrjardi, R., Raei, B., 2024. Monitoring and assessment of spatiotemporal soil salinization in the Lake Urmia region. Environmental Monitoring and Assessment 196, 958.
Mohammadzadeh, K., Bahmani, S., 2018. Estimation of water surface temperature using remote sensing techniques and GIS (case study: Lake Urmia), Conference on Civil Engineering, Architecture and Urban Development of Islamic World Countries, Tabriz. (in Persian)
Mohanty, V., Behera, D., Panda, A., Swetanisha, S., 2025. Comparative Study of ARIMA and Deep Learning for NDVI Forecasting Using Landsat 8 Data. Indian Journal of Science and Technology 18, 922–936.
Naji, M., Azizi, Q., Yousefi, P., 2024. Evaluation of water surface and vegetation cover changes in the Lake Urmia basin based on remote sensing data (1993 and 2023), Second Conference on Geography and Environmental Sustainability (Water Resources, Issues and Challenges in Iran), Kermanshah. (in Persian)
Priya, R., Vani, K., 2024. Vegetation change detection and recovery assessment based on post-fire satellite imagery using deep learning. Scientific Reports 14.
Rabiei, S., Babaeian, E., Grunwald, S., 2025. Deep Learning-Based Short- and Mid-Term Surface and Subsurface Soil Moisture Projections from Remote Sensing and Digital Soil Maps. Remote Sensing 17.
Rhif, M., Abbes, A., Martinez, B., Farah, I., 2020. Deep Learning Models Performance for NDVI Time Series Prediction: A Case Study on North West Tunisia. Conference: 2020 Mediterranean and Middle-East Geoscience and Remote Sensing Symposium (M2GARSS)
Rokni, K., Ahmad, A., Selamat, A., Hazini, S., 2014. Water Feature Extraction and Change Detection Using Multitemporal Landsat Imagery. Remote Sensing 2014, 4173–4189.
S, K., Mariem, S., Ghannem, S., Mouelhi, S., Ghazouani, H., Bechir, B.N., 2025. A framework based on mechanistic modelling and machine learning for soil moisture estimation. Discover Soil 2.
Shamloo, N., Sattari, M., Valizadeh Kamran, K., Aydin, H., 2023. Evaluation of agricultural drought composite index (CDI) prediction methods based on satellite imagery using deep learning and machine learning approaches. Water and Soil 37, 787–807. (in Persian)
Shuaiying, Z., An, W., Zhang, Y., Cui, L., Xie, C., 2022. Wetlands Classification Using Quad-Polarimetric Synthetic Aperture Radar Through Convolutional Neural Networks Based on Polarimetric Features. Remote Sensing 14, 5133.
Tahmouresi, M.S., Niksokhan, M.H., Ehsani, A.H., 2024. Enhancing spatial resolution of satellite soil moisture data through stacking ensemble learning techniques. Scientific Reports 14, 25454.
Wang, Y., Shi, L., Hu, Y., Hu, X., Song, W., Wang, L., 2024. A comprehensive study of deep learning for soil moisture prediction. Hydrology and Earth System Sciences 28, 917–943.
Younesi Sinki, A., Amini, J., 2024. Estimation of soil moisture using the integration of Sentinel-1 radar satellite data and ground-based soil moisture sensor data, 28th National Conference on Surveying Engineering and Geospatial Information (Geomatics 2024), Tehran. (in Persian)
Yu, W., Li, J., Liu, Q., Zhao, J., Dong, Y., Wang, C., Lin, S., Xinran, Z., Zhang, H., 2021. Spatial-Temporal Prediction of Vegetation Index with Deep Recurrent Neural Networks. IEEE Geoscience and Remote Sensing Letters PP, 1–5.
Yu, Y., Filippi, P., Bishop, T., 2025. Field-scale soil moisture estimated from Sentinel-1 SAR data using a knowledge-guided deep learning approach. Conference: IGARSS 2025 - 2025 IEEE International Geoscience and Remote Sensing Symposium
Zhang, K., Yu, K., Xu, N., Wang, C., Lin, Y., 2024. Machine Learning-Based Soil Moisture Retrieval Using CYGNSS and Interpolated SMAP Data. ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences X-4-2024, 447–452.