Downscaled ERA5 monthly precipitation data using Multi-Fidelity Gaussian Processes between 1980 and 2012 for the Upper Beas and Sutlej Basins, Himalayas
GB/NERC/BAS/PDC/01769
Abstract:
The dataset is the output of a statistical model which downscales ERA5 monthly precipitation data using gauge measurements from the Upper Beas and Sutlej Basins in the Western Himalayas. Multi-Fidelity Gaussian Processes (MFGPs) are used to generate more accurate precipitation values between 1980 and 2012, including over ungauged areas of the basins. MFGPs are a probabilistic machine learning method that provides principled uncertainty estimates via the prediction of probability distributions. These predictions can therefore be used to estimate the likelihood of extreme precipitation events which have led to droughts, floods, and landslides.
Funding from UK Engineering and Physical Sciences Research Council [grant number: 2270379].
Citation:
Tazi, K. (2023). Downscaled ERA5 monthly precipitation data using Multi-Fidelity Gaussian Processes between 1980 and 2012 for the Upper Beas and Sutlej Basins, Himalayas (Version 1.0) [Data set]. NERC EDS UK Polar Data Centre. https://doi.org/10.5285/b2099787-b57c-44ae-bf42-0d46d9ec87cc
Access Data
GET DATA
REFERENCE MATERIALS
License
| Access: | No restrictions apply. |
|---|---|
| Use: | Data supplied under Open Government Licence v3.0 http://www.nationalarchives.gov.uk/doc/open-government-licence/version/3/. |
Basic Information
| Creation Date: | 2023-08-09 |
|---|---|
| Dataset Progress: | Complete |
| Dataset Language: | English |
| ISO Topic Categories: |
|
| Parameters: |
|
Additional Information
| Quality: | Data is a model output so should be complete. NaNs in the netCDF file correspond to locations outside of the study areas. | |
|---|---|---|
| Lineage/Methodology: | The MFGP model is trained using all stations data from Bannister et al. and ERA5 data over the study area (30 degrees North - 33.5 degrees North, 75.5 degrees East - 83 degrees East) between 1980 and 2012. The model takes as inputs time, latitude, longitude, and elevation. The raw model outputs the predictions at ERA5 and gauge fidelity levels in the form of normal distributions. The distributions need to be transformed to get the monthly precipitation values in mm/day. The dataset includes the raw mean and variance of the model output for both fidelity levels. The dataset also includes the transformed mean and upper and lower 95 percent confidence interval bounds for the high-fidelity output. This corresponds to the accurately downscaled data. The user can transform the data on their own via inverse Box-Cox transformations using the scaling factors separately provided. | |
Extent
| Temporal Coverage: | |
|---|---|
| Start Date | 1980-01-01 |
| End Date | 2012-12-01 |
| Spatial Coverage: | |
| Latitude | |
| Southernmost | 30.28125 |
| Northernmost | 32.90625 |
| Longitude | |
| Westernmost | 75.84375 |
| Easternmost | 82.40625 |
| Altitude | |
| Min Altitude | 306 |
| Max Altitude | 6295 |
| Depth | |
| Min Depth | N/A |
| Max Depth | N/A |
| Data Resolution: | |
| Latitude Resolution | 0.0625 degrees |
| Longitude Resolution | 0.0625 degrees |
| Horizontal Resolution Range | N/A |
| Vertical Resolution | N/A |
| Vertical Resolution Range | N/A |
| Temporal Resolution | Monthly |
| Temporal Resolution Range | N/A |
| Location: | |
| Location | Asia |
| Detailed Location | Upper Beas Basin, Himalayas |
| Location | Asia |
| Detailed Location | Upper Sutlej Basin, Himalayas |
Instrumentation
| Data Collection: | Python v3.10 |
|---|
Storage
| Distribution: | |
|---|---|
| Distribution Media | Online Internet (HTTP) |
| Distribution Size | 224 MB |
| Distribution Format | N/A |
| Fees | N/A |
| Data Storage: | Three files: - lambdas_1980_2012.csv (3.9 kB) - mfgp_predictions_1980_2012.csv (114 MB) - mfgp_predictions_1980_2012.nc (111 MB) |