Remote sensing data from Earth orbiting satellites have become indispensable in modern geo-spatial sciences. The technologies underlying the capture of remote sensing data have evolved over the decades which have resulted in an improvement in the data quality, rate of availability and processing.
The workshop will cover tasks such as generating Land Surface Temperature (LST) product from satellite imagery from scratch, extraction of information from ready-made products and raster algebra. Participants will go through a workflow that will present itself as a solution to a real life problem. The main Python libraries or frameworks to be used are rasterio, earthpy, pandas, matplotlib and geopandas. The data to be used will be Landsat 8 satellite imagery.
The first part of the workflow focuses on the extraction of intermediate products that are useful for the calculation of LST from satellite imagery. These products are Normalized Difference Vegetation Index (NDVI), Land Surface Emissivity (LSE) and Fractional Vegetation Cover (FVC). These products are not only useful for calculation of LST but are applicable in other remote sensing applications such as vegetation health monitoring and land cover classification. This section will also equip participants with raster algebra skills using Python.
The second part will cover the pre-processing activity of correcting Landsat 8 thermal bands for the extraction of LST and ultimately generate the LST. The participants will learn how to perform other mathematical operations on raster data using Python.
Finally, LST values at certain desired locations will be extracted. This will equip participants with skills on how to extract information stored as raster files to point features using geospatial Python libraries. In all sections of the workshop, intermediate results will be visualized within the Jupyter Notebook to give participants a hands-on feel of visualization with Python.
It is expected that at the end of a successful completion of the workshop, participants will be able to generate LST from scratch using Landsat 8 imagery and by extension all Landsat imagery with thermal bands. Also, participants should be able to derive other useful products like NDVI from any remote sensing image using the appropriate data and finally acquire raster processing skills useful in other applications.