InVEST urban cooling model calibration

Overview

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.

Installation

This library requires the gdal library, which can easily be installed with conda as in:

conda install -c conda-forge gdal

Then, this library can be installed as in:

pip install invest-ucm-calibration

An alternative for the last step is to clone the repository and install it as in:

git clone https://github.com/martibosch/invest-ucm-calibration.git
python setup.py install

TODO

• 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)
• Test calibration based on cc_method='intensity'
• 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

Acknowledgments

• 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.