Scholarlink

<p>Correct retrieval of Sea Surface Temperature (SST) from different satellite radiometers is a very important task for many environmental issues including climate change. Little was done to establish the accuracy of determining the Caspian Sea SST using remote sensing data. For instance, for the Landsat program, this accuracy is unknown. The Landsat program was launched over 50&#xa0;years ago and till today this is an important source of remote sensing data, including SST, with high spatial resolution. In our previous work (Bocharov et al., 2024), we have assessed the accuracy of determining the SST using Landsat data with the help of <i>in-situ</i> measurements of water temperature in the Caspian Sea. These <i>in-situ</i> data were obtained from two drifter experiments conducted in 2006–2007 and 2008. In total, six Surface Velocity Program Barometer (SVPB) drifters were used in these experiments. Based on the results of the comparison between satellite and in-situ SST data, it was recommended to use Landsat Level-2 product data, and for the Caspian Sea, a correction of + 1.0&#xa0;°C was suggested. In this study, we examine seasonal variations in the accuracy of SST retrieval from Landsat Level-2 data. The results of the studies revealed that for the Caspian Sea in the summer months, Landsat Level-2 product data correspond well to the actual in situ temperature. When analyzing measurements with a difference of no more than 2&#xa0;h from the satellite overpass, the average SST according to Landsat Level-2 data was slightly lower than the actual SST by 0.04&#xa0;°C and in case of all measurements done during the day—by 0.06&#xa0;°C. In winter months, the SST according to Landsat Level-2 data also turned out to be lower than the actual SST. When analyzing measurements with a difference of no more than 2&#xa0;h from the satellite overpass, the error was 1.43&#xa0;°C and for all measurements during the day 1.24&#xa0;°C. It was found that the algorithm for SST retrieval from Landsat Level-2 works quite well for waters with a temperature of 12–22&#xa0;°C. For waters with a temperature above 22&#xa0;°C, the amplitude of deviations increases, but the average value corresponds well to the actual SST. For a cold sea with a temperature below 8&#xa0;°C, the SST according to Landsat Level-2 data was underestimated in most cases and had maximum deviations from drifter readings.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Interseasonal Changes in the Accuracy of Retrieval of Sea Surface Temperature of the Caspian Sea based on LANDSAT-5, -7 Satellite Data

  • 期刊论文

  • 发表日期：

    2025年7月12日

查看全文
  • Alexander V. Bocharov,
  • Andrey G. Kostianoy,
  • Sergey A. Lebedev

摘要

Correct retrieval of Sea Surface Temperature (SST) from different satellite radiometers is a very important task for many environmental issues including climate change. Little was done to establish the accuracy of determining the Caspian Sea SST using remote sensing data. For instance, for the Landsat program, this accuracy is unknown. The Landsat program was launched over 50 years ago and till today this is an important source of remote sensing data, including SST, with high spatial resolution. In our previous work (Bocharov et al., 2024), we have assessed the accuracy of determining the SST using Landsat data with the help of in-situ measurements of water temperature in the Caspian Sea. These in-situ data were obtained from two drifter experiments conducted in 2006–2007 and 2008. In total, six Surface Velocity Program Barometer (SVPB) drifters were used in these experiments. Based on the results of the comparison between satellite and in-situ SST data, it was recommended to use Landsat Level-2 product data, and for the Caspian Sea, a correction of + 1.0 °C was suggested. In this study, we examine seasonal variations in the accuracy of SST retrieval from Landsat Level-2 data. The results of the studies revealed that for the Caspian Sea in the summer months, Landsat Level-2 product data correspond well to the actual in situ temperature. When analyzing measurements with a difference of no more than 2 h from the satellite overpass, the average SST according to Landsat Level-2 data was slightly lower than the actual SST by 0.04 °C and in case of all measurements done during the day—by 0.06 °C. In winter months, the SST according to Landsat Level-2 data also turned out to be lower than the actual SST. When analyzing measurements with a difference of no more than 2 h from the satellite overpass, the error was 1.43 °C and for all measurements during the day 1.24 °C. It was found that the algorithm for SST retrieval from Landsat Level-2 works quite well for waters with a temperature of 12–22 °C. For waters with a temperature above 22 °C, the amplitude of deviations increases, but the average value corresponds well to the actual SST. For a cold sea with a temperature below 8 °C, the SST according to Landsat Level-2 data was underestimated in most cases and had maximum deviations from drifter readings.