This repository is a supplementary to the manuscript "Large model structural uncertainty in global projections of urban heat waves".
The objectives of this project are:
- Use extreme gradient boosting (XGBoost or XGB) to train the models (emulators) from the CESM-LENS (with urban specific variable) simulations
- Apply the models (emulators) to CMIP5 simulations to predict Urban daily maximum of average 2-m temperature, and project Global Urban Heat Waves (UHWs)
- Analysis the uncertainties in global projections of UHWs
- If you do not have the "conda" system
# Download and install conda
$ wget https://repo.continuum.io/miniconda/Miniconda3-latest-Linux-x86_64.sh
$ chmod +x Miniconda3-latest-Linux-x86_64.sh
$ ./Miniconda3-latest-Linux-x86_64.sh
# Edit .bash_profile or .bashrc
PATH=$PATH:$HOME/.local/bin:$HOME/bin:$HOME/miniconda3/bin
# Activate the conda system
$source .bash_profile
# OR source .bashrc- Create and activate your own conda environment
# Create an environment "partmc" and install the necessary packages
conda env create -f environment.yml
# Activate the "partmc" environment
conda activate uhws| Tasks | Folders | Fig or Tab in paper | Fig or Tab in preprint |
|---|---|---|---|
| data preparation | 1_data_prep | Tab S1, Fig S3 (extreme_range) | Tab 1 |
| model development | 2_model_dev* | ||
| model validation | 3_model_valid | Fig S4 (figures_rmse) | Fig 7 |
| model application | 4_model_app | ||
| data analysis | 5_event_analysis* | Fig 1 (urban_gridcell), Fig 2 (uhws), Fig 3 (uncertainty), Fig 4 (intensity), Fig S1 (SNR), Fig S2 (location), Fig S5 (warming), Fig S6 (uhws_min), Fig S7 (uncertainty_min), Fig S8 (intensity_min) | Fig 1 - Fig 6 |
-
1_data_prep
Num Folder Comments How to get it? 1.1 CESM-LE-members-csv* CESM large ensemble features and urban temperature Raw CESM data and scripts CESM_raw_nc_to_csv/*.py 1.2 CESM-LE-members-urban-temp-extracted-csv*/urban_heat_LE_*.csv CESM large ensemble urban temperature only (for convenience) Data 1.1 and scripts CESM_label_only_prep/*.py 1.3 ensem_training_data*/*.csv training data Data 1.1, Raw CESM data (for 2051-2080 yrs training data), and scripts CESM_training_data/* 1.4 CMIP5_tasmax_nc CMIP gridcell maximum temperature Download from website 1.5 CMIP5_tasmax_csv CMIP urban gridcell only maximum temperature (for calculating urban heat) Data 1.4 and scripts CMIP_gridcell_temp_prep/*.ipynb 1.6 CMIP5-RCP85_nc CMIP features Download from website 1.7 CMIP5-RCP85_csv CMIP urban gridcell only features (for predicting urban temperatures) Data 1.6 and scripts CMIP_feature_prep/*.ipynb 1.8 feature_dist_95 feature ranges of CMIP and CESM training data Data 1.7, Data 1.3, and script get_feature_extremes/get*.ipynb 1.9 CESM-LE-members-gridcell-temp-extracted-csv/TREFHTMX_heat_LE_*.csv CESM gridcell maximum temperature Raw CESM data and scripts CESM_gridcell_temp_prep/*.py -
2_model_dev*
Num Folder Comments How to get it? 2.1 lat_lon_dict.dat lat and lon pairs Data 1.1 and script get_lat_lon_dict_ls.ipynb 2.2 lat_ls.dat lat list (for distributing training) Data 1.1 and script get_lat_lon_dict_ls.ipynb 2.3 ensem_model* emulators from machine learning Data 1.3, Data 2.1, and Data 2.2, and scripts ens*.py and ens*.sub -
3_model_valid*
Num Folder Comments How to get it? 3.1 CESM_validation* CESM urban temperature predictions (for validation) Data 1.1, Data 2.1, Data 2.3, and scripts pred/* 3.2 model-validation* (available at data folder) rmse and pcc of CESM predictions Data 3.1 and script eval/model_evaluation.ipynb 3.3 model-validation-diff* (available at data folder) warming difference Data 3.1 and script eval/model_diff_evaluation.ipynb -
4_model_app*
Num Folder Comments How to get it? 4.1 CMIP5_pred* CMIP urban temperature predictions Data 1.7, Data 2.1, Data 2.3, and scripts -
5_event_analysis*
Num Folders Comments How to get it? 5.1 uhws*/UHWs_CMIP (available at data folder) urban heat waves from CMIP Data 4.1 and script _get_data_CMIP_2006_2061.ipynb 5.2 uhws/HWs_CMIP (available at data folder) background heat waves from CMIP Data 1.5 and script _get_data_CMIP_2006_2061_gridcell.ipynb 5.3 uhws*/UHWs_CESM (available at data folder) urban heat waves from CESM Data 1.2 and script _get_data_CESM-LE_2006_2061.ipynb 5.4 uhws/HWs_CESM (available at data folder) background heat waves from CESM Data 1.9 and script _get_data_CESM-LE_2006_2061_gridcell.ipynb
- We would like to acknowledge high-performance computing support from Cheyenne (doi:10.5065/D6RX99HX) provided by NCAR's Computational and Information Systems Laboratory, sponsored by the National Science Foundation.
- The CESM project is supported primarily by the National Science Foundation (NSF).
- This work is based upon material supported by the NCAR, which is a major facility sponsored by the NSF under Cooperative Agreement No. 1852977.
- We thank AWS for providing AWS Cloud Credits for Research.
- L.Z. acknowledges the financial support from the Start-up Grant from University of Illinois, Urbana-Champaign.