UHI Twin Cities

Calibration of the InVEST urban cooling model in Twin Cities, USA

Instructions to reproduce

1. Clone the repository

You can use git to clone this repository as in:

git clone https://github.com/martibosch/uhi-twin-cities.git

or download as a zip file and extract it.

2. Software requirements

First of all, conda is required to automatically install all the sofware dependencies. See its installation page and follow the steps to install it in your operating system.

Additionally, GNU Make is used to manage the execution of the calibration workflow, which is usually built-in with Linux and OSX systems. Windows users can install it from the Anaconda prompt as in:

# ACHTUNG: You only need to run this in Windows
conda install -c conda-forge make

Then, from the root of this repository, you can create a conda environment with all the required software dependencies as in:

make create_environment

and then activate it as in:

conda activate uhi-twin-cities

3. Data requirements

Copy the UCM_CalibrationData from the Google Drive to the data/raw directory of this repository so that the directory structure is of the form:

|─ data
|  └─ raw
|     └─ UCM_CalibrationData
|        |─ InVEST_Inputs
|        |─ Twine_UHI_2016
|        └─ LandSurfaceTemperature2016
|
|─ .gitignore
|─ LICENSE
|─ Makefile
|─ README.md
└─ environment.yml

4. Calibrate the model

The calibration of the urban cooling model for the July4-6_2012_DayTemp1.tif, July4-6_2012_NightTemp1.tif, JJA_Day_Temp1.tif, JJA_Night_Temp1.tif and lst2016_utm_c/hdr.adf reference temperature rasters can be executed as in:

make calibrate

which will dump the calibrated parameters for each file in the data/processed directory (which will be created automatically if it does not exist).

5. Generate the calibration reports

Reports for the calibration of the urban cooling model for each of the reference temperature rasters can be obtained as in:

make calibration_reports

which will generate a PDF calibration report and dump it to the reports directory (which will be created automatically if it does not exist).

Project based on the cookiecutter data science project template. #cookiecutterdatascience

martibosch / uhi-twin-cities