مقایسه عملکرد سامانه‌های google earth engine و snap در تخمین پارامترهای کیفی آب دریاچه سد میناب با استفاده از تصاویر ماهواره‌ای چندتفکیک‌مکانی

این پژوهش به مقایسه کارایی سامانه (gee)google earth engine و sentinel application platform (snap) در بررسی تغییرات زمانی کیفیت آب دریاچه سد میناب و ارتباط آن با متغیرهای مختلف محیطی شامل دمای آب (sst)، رسوبات معلق (ssc) و غلظت کلروفیل_a پرداخته است. داده‌های این پژوهش طی دوره زمانی 2016 تا 2023 جمع‌آوری و تحلیل شده‌اند. نتایج نشان داد که کیفیت آب دریاچه سد میناب طی دوره مورد بررسی روندی کاهشی داشته است. به‌طور خاص، غلظت کلروفیل_a در فصل‌های بهار و تابستان به دلیل افزایش sst افزایش یافته، در حالی که در پاییز و زمستان به دلیل کاهش دما و نور خورشید، کاهش یافته است. همچنین، تاثیر رسوبات معلق ssc بر کاهش کیفیت آب به ویژه در ماه‌های سردتر مشهود بوده است. مقایسه نتایج نشان داد، snap در تصحیح اتمسفری و استخراج جزئیات مکانی (با دقت 15-20% بالاتر) از gee عملکرد برتری دارد، در مقابل gee در پردازش داده‌های بلندمدت و کلان (کاهش 40% زمان پردازش) نسبت به snap کارآمدتر بود. snap تغییرات فصلی کلروفیل_a را با خطای31/4 شناسایی کرد، درحالی که این مقدار برای gee برابر با 44/4 براورد نمود. لذا snap برای مطالعات دقیق و محلی (با قابلیت تنظیم پارامترهای الگوریتمی) مناسب‌تر است در حالی‌که gee به دلیل دسترسی به داده‌های ابری و ابزارهای یادگیری ماشین، برای تحلیل‌های گسترده و سریع گزینه بهینه‌ای است. این تحقیق نشان داد که عوامل محیطی متعدد مانند دما، نور، مواد مغذی، و جریانات آبی تاثیرات پیچیده‌ای بر تولید فیتوپلانکتون‌ها و در نتیجه غلظت کلروفیل_a دارند و تنها نمی‌توان به یک عامل خاص برای توضیح تغییرات غلظت کلروفیل_a تکیه کرد. در نهایت، یافته‌ها نشان‌دهنده اهمیت کاربرد و تلفیق قابلیت‌های این دو سامانه (gee مقیاس‌پذیری و snap صحت) برای پایش و مدیریت منابع آب است.

comparative performance analysis of google earth engine and snap platforms for estimating water quality parameters in minab dam reservoir using multi-spectral resolution satellite imagery

abstractintroduction access to clean freshwater has become a critical global challenge due to pollution from anthropogenic activities like urbanization and intensive agriculture. these pressures have significantly degraded lake water quality, increasing the risk of harmful algal blooms—a phenomenon that is exacerbated by climate change. lakes, as multifunctional ecosystems supporting water supply, irrigation, and energy production, are particularly vulnerable to eutrophication. this process is primarily driven by excessive aquatic plant growth, quantified through chlorophyll-a (chl-a) concentrations. as the dominant photosynthetic pigment in phytoplankton, chl-a serves as a sensitive biomarker for nutrient loading and lake trophic status due to its rapid response to environmental shifts. water surface temperature (sst) and suspended sediment concentration (ssc) are equally vital water quality parameters. sst profoundly influences aquatic ecosystems’ biogeochemical cycles, while turbidity—a measure of light-scattering suspended particles—reflects sediment dynamics and anthropogenic pollution. remote sensing technologies, particularly satellite platforms like sentinel-2, have emerged as powerful tools for synoptic water quality monitoring. this study evaluates the efficacy of google earth engine (gee) and sentinel application platform (snap) in analyzing temporal variations of water quality parameters (chl-a, sst, ssc) in minab dam lake (2016–2023). the investigation further explores interrelationships between these variables and environmental drivers.materials and methods this study evaluated the water quality of minab esteghlal dam lake using remote sensing techniques, integrating field data from the hormozgan regional water department with multi-sensor satellite imagery. sentinel-2 msi level-2a and landsat-8 oli/tirs collection 2 level-2 data were acquired from the copernicus open access hub (2016–2023) to analyze chl-a, sst, and ssc. the normalized difference water index (ndwi) was applied to delineate water boundaries, followed by atmospheric correction of sentinel-2 data using sen2cor. the gee facilitated large-scale temporal processing, while the snap enabled advanced spectral analysis, including chl-a estimation via oc2/oc3 algorithms and ssc quantification using red-band reflectance. landsat-8 thermal bands were used to derive sst through radiative transfer equations. final outputs in tiff format were visualized in arcgis with standardized symbology, ensuring accurate spatial representation of water quality dynamics. this integrated approach leveraged cloud-based (gee) and desktop (snap/arcgis) platforms to optimize efficiency and precision in monitoring the lake’s ecological parameters.results and discussion the analysis demonstrated a declining trend in minab dam lake’s water quality during the study period. both snap and gee platforms consistently revealed seasonal chl-a dynamics, with concentrations peaking in spring/summer (attributed to elevated sst and optimal phytoplankton growth conditions) and declining in autumn/winter (due to reduced temperature and solar radiation). snap exhibited superior performance in atmospheric correction and spatial detail extraction, achieving 15-20% higher accuracy than gee in chl-a quantification, with lower estimation errors (rmse: 3.14 vs. gee’s 4.44). this precision stems from snap’s advanced algorithmic parameter customization capabilities, making it ideal for localized, high-resolution studies. conversely, gee excelled in large-scale, long-term trend analysis, processing big data 40% faster through its cloud-computing infrastructure and machine learning tools, albeit with marginally reduced precision.the sst-chl-a relationship exhibited nonlinear seasonality, with stronger correlations observed during warmer months. while ssc generally suppressed chl-a concentrations through light attenuation, episodic nutrient inputs from sediment loads occasionally triggered transient phytoplankton blooms. notably, ssc impacts were more pronounced during colder months, exacerbating water quality degradation. these findings highlight the platforms’ complementary strengths: snap for process-oriented studies requiring atmospheric precision, and gee for synoptic, time-series investigations. the results underscore the importance of platform selection based on study objectives—whether high-precision local analysis or watershed-scale temporal monitoring.conclusion this study demonstrates the effectiveness of remote sensing platforms in monitoring the water quality dynamics of minab dam lake, revealing a clear declining trend over the 2016–2023 period. the seasonal variability of chl-a concentrations—peaking in spring/summer due to warmer sst and favorable growth conditions, then declining in autumn/winter—was consistently captured by both snap and gee. however, their complementary strengths highlight the importance of platform selection based on research objectives: snap proved superior for localized, high-precision analysis, excelling in atmospheric correction (15–20% greater accuracy) and spatial detail extraction. its customizable algorithms make it ideal for mechanistic studies requiring fine-tuned parameter adjustments.gee enabled efficient large-scale and long-term assessments, reducing processing time by 40% through cloud-based workflows. while slightly less precise, its machine learning tools and data scalability are invaluable for trend analysis.the complex interplay between environmental drivers—such as the nonlinear sst-chl-a relationship and dual role of ssc as both light attenuators and nutrient sources—underscores the need for integrated monitoring approaches. these findings provide actionable insights for water resource management: snap can guide targeted mitigation of eutrophication hotspots, while gee supports system-wide policy decisions. to enhance the monitoring and management of minab dam lake, a hybrid approach combining high-precision snap analysis for localized eutrophication hotspots and gee-based large-scale trend assessment is recommended. targeted mitigation strategies, such as reducing agricultural runoff and controlling sediment inflow, should be prioritized in critical zones identified through snap’s detailed chl-a mapping. meanwhile, gee’s rapid processing capabilities can support real-time policy adjustments by tracking seasonal water quality variations across the entire watershed. additionally, integrating in-situ sensors with remote sensing data can improve predictive modeling, particularly under climate change-induced stressors. finally, stakeholder collaboration between environmental agencies, local communities, and policymakers is essential to implement sustainable water management practices, ensuring long-term ecological balance in the lake ecosystem.