Automated calibration of the InVEST urban cooling model with simulated annealing
Citation: Bosch, M., Locatelli, M., Hamel, P., Remme, R. P., Chenal, J., and Joost, S. 2021. "A spatially-explicit approach to simulate urban heat mitigation with InVEST (v3.8.0)". Geoscientific Model Development 14(6), 3521-3537. 10.5194/gmd-14-3521-2021
See the user guide for more information, or the lausanne-heat-islands repository for an example use of this library in an academic article.
The easiest way to install this library is using conda (or mamba), as in:
conda install -c conda-forge invest-ucm-calibrationwhich will install all the required dependencies including InVEST (minimum version 3.11.0). Otherwise, you can install the library with pip provided that all the dependencies (including GDAL) are installed.
- Allow a sequence of LULC rasters (although this would require an explicit mapping of each LULC/evapotranspiration/temperature raster or station measurement to a specific date)
- Support spatio-temporal datasets with xarray to avoid passing many separate rasters (and map each raster to a date more consistently)
- Read both station measurements and station locations as a single geo-data frame
- The calibration procedure is based simulated annealing implementation of perrygeo/simanneal
- With the support of the École Polytechnique Fédérale de Lausanne (EPFL)
- This package was created with the ppw tool. For more information, please visit the project page.