PR for Urban Heat Island computation through OpenEO UDP

[adriaanzico]

This PR is to add a new algorithm into the APEx algorithm catalogue from Compass Informatics. This OpenEO public UDP uses Landsat 8/9 thermal imagery, and ESA CCI land cover maps to compute the urban heat island (UHI) effect for any urban AOI during the summer months of any given years. It uses https://openeo.dataspace.copernicus.eu/openeo/1.2 as backend.

The process is described below:

Algorithm:

1. Load Landsat TIRS B10, convert to °C.
2. Aggregate to JJA (June-August) mean across all years in the requested temporal extent.
3. Load ESA CCI land cover; resample to LST grid.
4. Separate urban (CCI=190) and rural pixels.
5. Compute local rural mean LST with a moving-window kernel.
6. UHI = LST_urban - local_rural_mean (urban pixels only).

The UDP was created with a Python script (algorithm_catalog/compass_informatics/openeo_udp/UHI_udp.py). It includes a summary and a description. It is named 'UHI_per_pixel' and can be found in namespace https://openeo.dataspace.copernicus.eu/openeo/1.2/processes/u:776bee94-53de-49f5-88d6-c3660d927048/UHI_per_pixel

The UDP has been tested in Python script algorithm_catalog/compass_informatics/openeo_udp/UHI_udp_run.py. The required inputs for this UDP are spatial_extent and temporal_extent. The output is a GeoTIFF showing the UHI value per pixel.

[JanssenBrm]

Thank you for this contribution @adriaanzico. I have created a new PR that allows me to do some quick fixes to the record as I am not allowed to push to the remote fork: #405. In this PR you can also preview the record, how it will show up in the APEx Algorithm Catalogue.

I'll close this one in favor of the other.