A GIS processing and visualization tool
poolsville_hacks 2022 Best Overall winner! 🎉🎉🎉
See how a world of your own making, or Earth in the future, or your favorite fictional world, would look like from space!
ClimateCNN uses a neural network to predict how a certain area on an Earthlike planet might look like given the area's topographic, temperature, and precipitation data.
Download this repository or run
git clone https://github.com/0something0/ClimateCNNCurrently, ClimateCNN requires Conda for its Rasterio and GDAL dependencies.
After, run
pip install -r requirements.txtin the project directory to download the dependencies.
ClimateCNN uses GeoTIFF files for rainfall, temperature, and heightmap data, and the PNG files for colormaps.
Currently, the project only supports inputs of one file per folder.
This is not needed if you are merely using the provided neural network model, rather than training your own.
If you are training your own model, you should provide every folder with the corrosponding data.
├── training_dataset
│ ├── color
│ ├── elevation
│ ├── rainfall
│ └── temperature
Only the elevation, rainfall, and temperature folders should be given input files. Color_output will output prediction data in chunks.
.
├── prediction_dataset
│ ├── color_output
│ ├── elevation
│ ├── rainfall
│ └── temperature
Single, contigious files should be provided for all inputs.
- Remove GDAL dependencies
- Multiple file input support
- Experiment with different parameters
- Interactivity
Amante, C. ETOPO1 1 Arc-Minute Global Relief Model: Procedures, Data Sources and Analysis. National Geophysical Data Center, NOAA, 2009. DOI.org (Datacite), https://doi.org/10.7289/V5C8276M.
Fick, S.E. and R.J. Hijmans, 2017. WorldClim 2: new 1km spatial resolution climate surfaces for global land areas. International Journal of Climatology 37 (12): 4302-4315.
NASA Visible Earth - Home. https://visibleearth.nasa.gov/collection/1484/blue-marble.