Thermal sharpening of Sentinel-3 pictures: From 1 Km to 10m utilizing Python in Google Colab

1. 🌅 Introduction
2. 💾 Downloading Sentinel-3 (1000 m) and Sentinel-2 pictures (10 m)
3. ⚙️ Sentinel-3 Picture Processing
4. 🌡️ Temperature-NDVI House
5. 📐 Sharpening the Thermal Picture (1000 m to 10 m)
6. 🗺️ Visualization of the Sharpened Thermal Picture
7. 📄 Conclusion
8. 📚 References

🌅 Introduction

Downscaling the thermal imagery captured by satellites has been extensively studied as a result of tradeoff between the spatial and temporal decision of satellites that present thermal pictures. For instance, the revisit cycle of Landsat-8 is 16 days, with an authentic thermal decision of 100 meters. In distinction, Sentinel-3 can present day by day thermal pictures, however at a spatial decision of 1000 meters.

One strategy to deal with the coarse decision of thermal pictures might be launching extra satellites outfitted with thermal sensors, equivalent to NASA’s Landsat-9, launched in September 2021. On this case, the temporal decision for each Landsat-8 and Landsat-9 is 8 days (as an alternative of 16 days with one satellite tv for pc), assuming clear skies.

Nevertheless, as you may guess, this strategy requires a multimillion-dollar funding and several other years of effort. As a substitute, researchers have centered on statistical strategies, correlating the seen/near-infrared (VNIR) bands from satellites with greater spatial decision (however decrease temporal decision) to thermal pictures from satellites with decrease spatial decision (however greater temporal decision). For instance, research have proven that the Normalized Distinction Vegetation Index (NDVI) calculated from VNIR bands of Sentinel-2 (10m, each 5 days) may be inversely correlated with thermal pictures from Sentinel-3 (1000m, day by day).

However how can we use this…