Multivariate bias correction of ERA5 climatic data for assessing climate-related road vulnerabilities in Nigeria
Abstract The vulnerability of roads to climate change is a significant challenge in Nigeria, exacerbated by the need for high-resolution and comprehensive climatic data amid climatic data scarcity. This study addresses this issue by applying multivariate bias correction to the fifth generation of at...
Saved in:
| Main Authors: | , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Springer
2025-04-01
|
| Series: | Discover Geoscience |
| Subjects: | |
| Online Access: | https://doi.org/10.1007/s44288-025-00151-4 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Abstract The vulnerability of roads to climate change is a significant challenge in Nigeria, exacerbated by the need for high-resolution and comprehensive climatic data amid climatic data scarcity. This study addresses this issue by applying multivariate bias correction to the fifth generation of atmospheric reanalysis of the global (ERA5) climatic data from the European Centre for Medium-Range Weather Forecasts (ECMWF), supplementing existing data for road vulnerability assessment. Daily rainfall, minimum, and maximum temperatures from forty weather stations of the Nigerian Meteorological Station were collected, alongside ERA5 data at a 0.25° resolution from 1980 to 2015. The study employs multivariate bias correction using both N-dimension probability (MBCn) and ranked correlation dependence structures (MBCr) to assess the replicability of ERA5. Evaluation metrics such as mean absolute error (MAE), root mean square error (RMSE), and spatial mean difference are utilised to assess the bias correction techniques. Results reveal that ERA5 rainfall exhibits higher uncertainty than temperature, particularly in some parts of the Guinea savannah and swamp forest. The MBCn outperformed MBCr and MBCp in replicating minimum temperature (RMSE: 0.65–1.66 and MAE: 0.83–4.15) maximum temperature (RMSE: 0.76–1.55 and MAE 1.08–3.69) and rainfall (RMSE: 0.93–5.28 and MAE: 1.56–32.94). The study concludes that multivariate bias correction should be conducted before applying any reanalysis and model data as to enhance data reliability and support sustainable planning and adaptation strategies in transportation and infrastructure. |
|---|---|
| ISSN: | 2948-1589 |