Deep-learning-based harmonization and super-resolution of near-surface air temperature from CMIP6 models (1850–2100)
(2023) In International Journal of Climatology 43(3). p.1461-1479- Abstract
Future global temperature change will have significant effects on society and ecosystems. Earth system models (ESM) are the primary tools to explore future climate change. However, ESMs have great uncertainty and often run at a coarse spatial resolution (usually about 2°). Accurate high-spatial-resolution temperature dataset are needed to improve our understanding of temperature variations and for many other applications. We apply Super resolution (SR) in computer vision using deep learning (DL) to merge 31 ESMs data. The proposed method performs data merging, bias-correction, and spatial downscaling simultaneously. The CRU TS (Climate Research Unit gridded Time Series) data is used as reference data in the model training process. To... (More)
Future global temperature change will have significant effects on society and ecosystems. Earth system models (ESM) are the primary tools to explore future climate change. However, ESMs have great uncertainty and often run at a coarse spatial resolution (usually about 2°). Accurate high-spatial-resolution temperature dataset are needed to improve our understanding of temperature variations and for many other applications. We apply Super resolution (SR) in computer vision using deep learning (DL) to merge 31 ESMs data. The proposed method performs data merging, bias-correction, and spatial downscaling simultaneously. The CRU TS (Climate Research Unit gridded Time Series) data is used as reference data in the model training process. To find a suitable DL method, we select five SR methodologies with different structures. We compare the performances of the methods based on mean square error (MSE), mean absolute error (MAE) and Pearson correlation coefficient (R). The best method is used to merge the projected monthly data (1850–1900), and monthly future scenarios data (2015–2100), at the high spatial resolution of 0.5°. Results show that the merged data have considerably improved performance compared with individual ESM data and their ensemble mean (EM), both spatially and temporally. The MAE shows significant improvement; the spatial distribution of the MAE widens along the latitudes in the Northern Hemisphere. The MAE of merged data is ranging from 0.60 to 1.50, the South American (SA) has the lowest error and the Europe has the highest error. The merged product has excellent performance when the observation data is smooth with few fluctuations in the time series. This work demonstrates the applicability and effectiveness of the DL methods in data merging, bias-correction and spatial downscaling when enough training data are available. Data can be accessed at https://doi.org/10.5281/zenodo.5746632.
(Less)
- author
- Wei, Xikun ; Wang, Guojie ; Feng, Donghan ; Duan, Zheng LU ; Hagan, Daniel Fiifi Tawia ; Tao, Liangliang ; Miao, Lijuan ; Su, Buda and Jiang, Tong
- organization
- publishing date
- 2023
- type
- Contribution to journal
- publication status
- published
- subject
- keywords
- climate change, CMIP6, data-merging, deep-learning, temperature
- in
- International Journal of Climatology
- volume
- 43
- issue
- 3
- pages
- 1461 - 1479
- publisher
- John Wiley & Sons Inc.
- external identifiers
-
- scopus:85144222592
- ISSN
- 0899-8418
- DOI
- 10.1002/joc.7926
- language
- English
- LU publication?
- yes
- id
- 5dbc3127-ad63-4db0-85fb-486bd373548e
- date added to LUP
- 2023-01-23 15:34:56
- date last changed
- 2023-10-26 14:53:27
@article{5dbc3127-ad63-4db0-85fb-486bd373548e,
abstract = {{<p>Future global temperature change will have significant effects on society and ecosystems. Earth system models (ESM) are the primary tools to explore future climate change. However, ESMs have great uncertainty and often run at a coarse spatial resolution (usually about 2°). Accurate high-spatial-resolution temperature dataset are needed to improve our understanding of temperature variations and for many other applications. We apply Super resolution (SR) in computer vision using deep learning (DL) to merge 31 ESMs data. The proposed method performs data merging, bias-correction, and spatial downscaling simultaneously. The CRU TS (Climate Research Unit gridded Time Series) data is used as reference data in the model training process. To find a suitable DL method, we select five SR methodologies with different structures. We compare the performances of the methods based on mean square error (MSE), mean absolute error (MAE) and Pearson correlation coefficient (R). The best method is used to merge the projected monthly data (1850–1900), and monthly future scenarios data (2015–2100), at the high spatial resolution of 0.5°. Results show that the merged data have considerably improved performance compared with individual ESM data and their ensemble mean (EM), both spatially and temporally. The MAE shows significant improvement; the spatial distribution of the MAE widens along the latitudes in the Northern Hemisphere. The MAE of merged data is ranging from 0.60 to 1.50, the South American (SA) has the lowest error and the Europe has the highest error. The merged product has excellent performance when the observation data is smooth with few fluctuations in the time series. This work demonstrates the applicability and effectiveness of the DL methods in data merging, bias-correction and spatial downscaling when enough training data are available. Data can be accessed at https://doi.org/10.5281/zenodo.5746632.</p>}},
author = {{Wei, Xikun and Wang, Guojie and Feng, Donghan and Duan, Zheng and Hagan, Daniel Fiifi Tawia and Tao, Liangliang and Miao, Lijuan and Su, Buda and Jiang, Tong}},
issn = {{0899-8418}},
keywords = {{climate change; CMIP6; data-merging; deep-learning; temperature}},
language = {{eng}},
number = {{3}},
pages = {{1461--1479}},
publisher = {{John Wiley & Sons Inc.}},
series = {{International Journal of Climatology}},
title = {{Deep-learning-based harmonization and super-resolution of near-surface air temperature from CMIP6 models (1850–2100)}},
url = {{http://dx.doi.org/10.1002/joc.7926}},
doi = {{10.1002/joc.7926}},
volume = {{43}},
year = {{2023}},
}