Find research datasets worth reusing

Weather dataset (Typical Downscaled Year, Extreme Cold Year, Extreme Warm Year) for building energy simulations (.epw format) in future climate (2069-2098, RCP 8.5), Uccle KMI, Belgium

Typical Downscaled Year (TDY), Extreme Cold Year and Extreme Warm Year based on the methodology of Nik (2016), is extracted for the location of Uccle KMI (50°47'49"N, 4°21'29" E) from the EC-Earth driven convection-permitting climate model COSMO-CLM for the Belgian domain extended with land-surface scheme TERRA_URB(v2.0) making use of the SURY (Semi-empirical URban canopY) parameterization ( Wouters et al. 2016).  The integrations are identical to the ones which are described in Vanden Broucke et al. (2019). The climate model has a spatial resolution of 2.8 km and an hourly temporal resolution and is available for the recent past (1976-2004) and future (2070-2098, RCP 8.5 climate change scenario) as 30-year datasets. For this dataset, the TDY, ECY, and EWY are extracted for the future period. A bias correction is applied for the following variables: temperature (as described in Ramon et al. 2020), solar radiation and relative humidity as described in Ramon et al. (202X).

Weather dataset (Typical Downscaled Year, Extreme Cold Year, Extreme Warm Year) for building energy simulations (.epw format) in future climate (2069-2098, RCP 8.5), Leuven City centre, Belgium

Typical Downscaled Year (TDY), Extreme Cold Year and Extreme Warm Year based on the methodology of Nik (2016), is extracted for the location of Leuven City Centre (50°52'48"N 4°42'0" E) from the EC-Earth driven convection-permitting climate model COSMO-CLM for the Belgian domain extended with land-surface scheme TERRA_URB(v2.0) making use of the SURY (Semi-empirical URban canopY) parameterization ( Wouters et al. 2016).  The integrations are identical to the ones which are described in Vanden Broucke et al. (2019). The climate model has a spatial resolution of 2.8 km and an hourly temporal resolution and is available for the recent past (1976-2004) and future (2070-2098, RCP 8.5 climate change scenario) as 30-year datasets. For this dataset, the TDY, ECY, and EWY are extracted for the future period. A bias correction is applied for the following variables: temperature (as described in Ramon et al. 2020), solar radiation and relative humidity as described in Ramon et al. (202X).

Weather dataset (Typical Downscaled Year, Extreme Cold Year, Extreme Warm Year) for building energy simulations (.epw format) in future climate (2069-2098, RCP 8.5), Leuven Casa Blanca, Belgium

Typical Downscaled Year (TDY), Extreme Cold Year and Extreme Warm Year based on the methodology of Nik (2016), is extracted for the location of Casa Blanca neighbourhood Leuven (50°52'48"N, 4°43'48"E) from the EC-Earth driven convection-permitting climate model COSMO-CLM for the Belgian domain extended with land-surface scheme TERRA_URB(v2.0) making use of the SURY (Semi-empirical URban canopY) parameterization ( Wouters et al. 2016).  The integrations are identical to the ones which are described in Vanden Broucke et al. (2019). The climate model has a spatial resolution of 2.8 km and an hourly temporal resolution and is available for the recent past (1976-2004) and future (2070-2098, RCP 8.5 climate change scenario) as 30-year datasets. For this dataset, the TDY, ECY, and EWY are extracted for the future period. A bias correction is applied for the following variables: temperature (as described in Ramon et al. 2020), solar radiation and relative humidity as described in Ramon et al. (202X).

Weather dataset (Typical Downscaled Year, Extreme Cold Year, Extreme Warm Year) for building energy simulations (.epw format) in recent past climate, Leuven Casa Blanca, Belgium

Typical Downscaled Year (TDY), Extreme Cold Year and Extreme Warm Year based on the methodology of Nik (2016), is extracted for the location of the Casa Blanca Neighbourhood Leuven (50°52'48"N 4°43'48"E) from the EC-Earth driven convection-permitting climate model COSMO-CLM for the Belgian domain extended with land-surface scheme TERRA_URB(v2.0) making use of the SURY (Semi-empirical URban canopY) parameterization ( Wouters et al. 2016).  The integrations are identical to the ones which are described in Vanden Broucke et al. (2019). The climate model has a spatial resolution of 2.8 km and an hourly temporal resolution and is available for the recent past (1976-2004) and future (2070-2098) as 30-year datasets. For this dataset, the TDY, ECY, and EWY are extracted for the recent past period. A bias correction is applied for the following variables: temperature (as described in Ramon et al. 2020), solar radiation and relative humidity as described in Ramon et al. (202X).

Weather dataset (Typical Downscaled Year, Extreme Cold Year, Extreme Warm Year) for building energy simulations (.epw format) in recent past climate, Uccle KMI, Belgium

Typical Downscaled Year (TDY), Extreme Cold Year and Extreme Warm Year based on the methodology of Nik (2016), is extracted for the location of Uccle KMI (50°47'49"N 4°21'29" E) from the EC-Earth driven convection-permitting climate model COSMO-CLM for the Belgian domain extended with land-surface scheme TERRA_URB(v2.0) making use of the SURY (Semi-empirical URban canopY) parameterization ( Wouters et al. 2016).  The integrations are identical to the ones which are described in Vanden Broucke et al. (2019). The climate model has a spatial resolution of 2.8 km and an hourly temporal resolution and is available for the recent past (1976-2004) and future (2070-2098) as 30-year datasets. For this dataset, the TDY, ECY, and EWY are extracted for the recent past period. A bias correction is applied for the following variables: temperature (as described in Ramon et al. 2020), solar radiation and relative humidity as described in Ramon et al. (202X).

Weather dataset (Typical Downscaled Year, Extreme Cold Year, Extreme Warm Year) for building energy simulations (.epw format) in recent past climate, Leuven City centre, Belgium

Typical Downscaled Year (TDY), Extreme Cold Year and Extreme Warm Year based on the methodology of Nik (2016), is extracted for the location of city centre of Leuven (50°52'48"N, 4°42'0"E) from the EC-Earth driven convection-permitting climate model COSMO-CLM for the Belgian domain extended with land-surface scheme TERRA_URB(v2.0) making use of the SURY (Semi-empirical URban canopY) parameterization ( Wouters et al. 2016).  The integrations are identical to the ones which are described in Vanden Broucke et al. (2019). The climate model has a spatial resolution of 2.8 km and an hourly temporal resolution and is available for the recent past (1976-2004) and future (2070-2098) as 30-year datasets. For this dataset, the TDY, ECY, and EWY are extracted for the recent past period. A bias correction is applied for the following variables: temperature (as described in Ramon et al. 2020), solar radiation and relative humidity as described in Ramon et al. (202X).

opencc-by-4.0Sep 2024View details →

Dataset for "On the Extrapolation of Generative Adversarial Networks for downscaling precipitation extremes in warmer climates"

<h1>Code and Dataset for "On the Extrapolation of Generative Adversarial Networks for downscaling precipitation extremes in warmer climates"</h1> This dataset accompanies the research paper titled "On the Extrapolation of Generative Adversarial Networks for downscaling precipitation extremes in warmer climates", currently under review for the AGU Journal GRL. The study introduces a novel Regional Climate Model (RCM) emulator focusing on high-resolution climate downscaling for the New Zealand region. For additional insights and access to the codebase utilized in this research, please refer to our <a href="https://github.com/nram812/On-the-Extrapolation-of-Generative-Adversarial-Networks-for-downscaling-precipitation-extremes">Github Repository</a>. The code can also be found as a ".zip" file: *On-the-Extrapolation-of-Generative-Adversarial-Networks-for-downscaling-precipitation-extremes-main.  <h2>Aims</h2> Our study focuses on two important gaps in the literature regarding the extrapolation of empirical downscaling algorithms. First, we examine how well relationships learned from a historical period extrapolate to future unobserved climates. We compare two widely used algorithms, a GAN and a deterministic CNN baseline, that use a similar architecture (i.e. convolutional layers) trained in a model-as-truth framework to downscale daily precipitation over New Zealand. We evaluate their accuracy in capturing climate change signals in mean and extreme precipitation. Second, we explore whether training on future vs. only historical periods combined with different-sized training datasets can improve extrapolation skill.  <h2>Geographic Focus</h2> Our research focuses only on the New Zealand Region (165°E-184°W, 33°S-51°S).   <h2>Data Overview</h2> <h3>Training and Evaluation Data</h3> The training data used in this study (for our RCM emulator) spans the historical period and future period (SSP370) of simulation. It comprises daily accumulated precipitation as the primary target variable, alongside large-scale predictor variables.  <ul> <li> Resolution: The target variable is presented at a 12km resolution, reflecting the highest resolution face of RCM for the New Zealand region. Predictor variables are coarsened to a 1.5-degree resolution from original CCAM outputs using conservative interpolation.  </li> <li> Period Coverage: <ul> <li>Training Data: 1960-2100 (Depending on Experiment, see Table 1 for list of experiment configurations)</li> <li>Validation Data: 1985-2014 + 2070-2099 (to compute the climate change signal)</li> </ul> </li> <li> Models: <ul> <li>Training on: ACCESS-CM2</li> <li>Validated on: EC-Earth3, NorESM2-MM, CNRM-CM6-1, AWI-MR-1 </li> </ul> </li> </ul> <h3>File Structure</h3> <ul> <li> Training Data: <ul> <li>Target/Ground Truth (Y): <code>target_ACCESS-CM2_hist_ssp370_pr.nc</code></li> <li>Predictor (X): <code>predictor_ACCESS-CM2_hist_ssp370.nc</code></li> </ul> </li> <li> Evaluation Data: All other GCMs can be accessed in one single file, predictor and target variables have the dimensions (time, lat, lon, GCM). <ul> <li>Target/Ground Truth (Y): <code>Other_GCMs_hist_SSP370_target_fields_pr.nc</code></li> <li>Predictor (X): <code>Other_GCMs_hist_SSP370_predictor_fields.nc</code></li> </ul> </li> </ul> <h2>Methodological Insights</h2> <ul> <li> Regional Climate Model, Our Regional Climate Model training data is from the Conformal Cubic Atmospheric Model (CCAM) which is a global non-hydrostatic atmospheric model renowned for its variable-resolution cubic grid. . For more information about CCAM, please see the following <a href="https://agupubs.onlinelibrary.wiley.com/doi/abs/10.1029/2023JD038530">paper</a>. </li> <li> Predictor and Target Variables: Daily-averaged large-scale prognostic variables, including zonal wind, meridional wind, temperature, and specific humidity, are employed as predictors at the 500mb and 850mb pressure levels. These are normalized (see the GitHub repository for the mean and standard deviation fields). Precipitation is taken as is from CCAM and accumulated for each given day. Static predictors are also used in our model, which is stored in a GitHub repository. </li> <li> Training Framework: Our dataset benefits from the "perfect framework" training strategy, which uses CCAM-coarsened predictor variables. For more information about the perfect and imperfect training frameworks, see the following <a title="review" href="https://journals.ametsoc.org/view/journals/aies/3/2/AIES-D-23-0066.1.xml">review</a> </li> </ul> <table> <tbody> <tr> <td> Algorithm </td> <td> Training Data </td> <td> Period </td> </tr> <tr> <td> Deterministic Baseline </td> <td> Historical </td> <td> 1960-2014 (~21,000 days) </td> </tr> <tr> <td> Deterministic Baseline </td> <td> Future (SSP370) </td> <td> 2044-2099 (~21,000 days) </td> </tr> <tr> <td> Deterministic Baseline </td> <td> Historical and Future (SSP370) </td> <td> 1960-2099 (~51,000 days) </td> </tr> <tr> <td> Residual GAN </td> <td> Historical </td> <td> 1960-2014 </td> </tr> <tr> <td> Residual GAN </td> <td> Future (SSP370) </td> <td> 2044-2099 </td> </tr> <tr> <td> Residual GAN </td> <td> Historical and Future (SSP370) </td> <td> 1960-2099 </td> </tr> </tbody> </table> Table 1: The six RCM emulator experiments performed in this study.

opencc-by-4.0Nov 2022View details →

Downscaling climatic data 2021-2100 in North Korea

Please refer to the 'Readme' file.

opencc-by-4.0Mar 2023View details →

Supporting material for PyESDv1.0.1 An open-source Python framework for empirical-statistical downscaling of climate information

The nature and severity of climate change impacts varies significantly from region to region. Consequently, high-resolution climate information is needed for meaningful impact assessments and the design of mitigation strategies. This demand has led to an increase in the coupling of Empirical Statistical Downscaling (ESD) models to General Circulation Model (GCM) simulations of future climate. Here, we present a new open-source Python package (pyESD; github.com/Dan-Boat/PyESD) that implements several Perfect Prognosis ESD (PP-ESD) methods and the whole downscaling cycle. The latter includes routines for data preparation, predictor selection and construction, model selection and training, evaluation, utility tools for relevant statistical tests, visualization, and more. The package includes a collection of well-established Machine Learning algorithms and allows the user to choose a variety of estimators, cross-validation schemes, objective function measures, hyperparameter optimization, etc., in relatively few lines of code. The package is highly modular and flexible and allows quick and reproducible downscaling of any climate information, such as precipitation, temperature, wind speed, or even glacial retreat. The dataset presented here serves as supporting material for the package description and evaluation manuscript

opencc-by-4.0Apr 2023View details →

Deep Learning Regional Climate Model Emulators: a comparison of two downscaling training frameworks [datasets]

Outputs used in: van der Meer, M., de Roda Husman, S., Lhermitte, S.: Deep Learning Regional Climate Model Emulators: a comparison of two downscaling training frameworks <ul> <li>MAR(ACCESS1-3)_monthly_SMB.nc: MAR outputs with monthly values of SMB and components over the Antarctic ice sheet (1980--2100)</li> <li>MAR(ACCESS1-3)-stereographic_monthly_GCM_like.nc: MAR outputs upscaled to GCM resolution (1980--2100)</li> <li>ACCESS1-3-stereographic_monthly_cleaned.nc: GCM monthly outputs over the Antarctic ice sheet (1980--2100)</li> </ul> The up-to-date working versions of our experiments and source code can be found and are available on our GitHub: <a href="https://github.com/marvande/RCM-Emulator">https://github.com/marvande/RCM-Emulator</a> and at this link: <a href="https://doi.org/10.5281/zenodo.7875967">https://doi.org/10.5281/zenodo.7875967</a> Data usage notice: If you use any of these results, please acknowledge the work of the people involved in producing them. You should also refer to and cite the following paper: Cite as: Marijn van der Meer, Sophie de Roda Husman, S Lhermitte. Deep Learning Regional Climate Model Emulators: a comparison of two downscaling training frameworks. Authorea. December 27, 2022  DOI: <a href="https://doi.org/10.22541/essoar.167214210.02213149/v1">10.22541/essoar.167214210.02213149/v1</a>

High resolution downscaling of CMIP6 Earth System and Global Climate Models using deep learning for Iberia Part 8

Future projections of precipitation by the BM10 model forced by the seven GCMs used in the GMD paper "High resolution downscaling of CMIP6 Earth System and Global Climate Models using deep learning for Iberia".

High resolution downscaling of CMIP6 Earth System and Global Climate Models using deep learning for Iberia Part 5

Future projections of precipitation by the BM1 model forced by the seven GCMs used in the GMD paper "High resolution downscaling of CMIP6 Earth System and Global Climate Models using deep learning for Iberia".

High resolution downscaling of CMIP6 Earth System and Global Climate Models using deep learning for Iberia Part 3

Future projections of 2-meter minimum temperature by the CNN models (BM1, BM10 and BMdense) forced by the seven GCMs used in the GMD paper "High resolution downscaling of CMIP6 Earth System and Global Climate Models using deep learning for Iberia".

High resolution downscaling of CMIP6 Earth System and Global Climate Models using deep learning for Iberia Part 2

Future projections of 2-meter maximum temperature by the CNN models (BM1, BM10 and BMdense) forced by the seven GCMs used in the GMD paper "High resolution downscaling of CMIP6 Earth System and Global Climate Models using deep learning for Iberia".

High resolution downscaling of CMIP6 Earth System and Global Climate Models using deep learning for Iberia Part 11

Future projections of 2-meter maximum, mean and minimum temperatures by the BMlinear model forced by the seven GCMs used in the GMD paper "High resolution downscaling of CMIP6 Earth System and Global Climate Models using deep learning for Iberia".

High resolution downscaling of CMIP6 Earth System and Global Climate Models using deep learning for Iberia Part 1

Input data (ERA5 and the seven GCMs) used to train the CNN models (BMlinear, BM1, BM10 and BMdense) used in the GMD paper "High resolution downscaling of CMIP6 Earth System and Global Climate Models using deep learning for Iberia".

High resolution downscaling of CMIP6 Earth System and Global Climate Models using deep learning for Iberia Part 9

Historical projections of all predictands (2-meter maximum, mean and minimum temperatures, and precipitation) by all CNN models (BMlinear, BM1, BM10, BMdense) forced by the seven GCMs used in the GMD paper "High resolution downscaling of CMIP6 Earth System and Global Climate Models using deep learning for Iberia". Each CNN model architecture is available in the file "model.json" and its optimized weights for each case are available in the file "model_weights.h5".

High resolution downscaling of CMIP6 Earth System and Global Climate Models using deep learning for Iberia Part 7

Future projections of precipitation by the BMdense model forced by the seven GCMs used in the GMD paper "High resolution downscaling of CMIP6 Earth System and Global Climate Models using deep learning for Iberia".