Spatial Modeling of Precipitation Based on Data-Driven Warping of Gaussian Processes
Modeling and forecasting spatiotemporal patterns of precipitation is crucial for managing water resources and mitigating water-related hazards. Globally valid spatiotemporal models of precipitation are not available. This is due to the intermittent nature, non-Gaussian distribution, and complex geog...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-02-01
|
| Series: | Entropy |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1099-4300/24/3/321 |
| _version_ | 1827629062200229888 |
|---|---|
| author | Vasiliki D. Agou Andrew Pavlides Dionissios T. Hristopulos |
| author_facet | Vasiliki D. Agou Andrew Pavlides Dionissios T. Hristopulos |
| author_sort | Vasiliki D. Agou |
| collection | DOAJ |
| description | Modeling and forecasting spatiotemporal patterns of precipitation is crucial for managing water resources and mitigating water-related hazards. Globally valid spatiotemporal models of precipitation are not available. This is due to the intermittent nature, non-Gaussian distribution, and complex geographical dependence of precipitation processes. Herein we propose a data-driven model of precipitation amount which employs a novel, data-driven (non-parametric) implementation of warped Gaussian processes. We investigate the proposed warped Gaussian process regression (wGPR) using (i) a synthetic test function contaminated with non-Gaussian noise and (ii) a reanalysis dataset of monthly precipitation from the Mediterranean island of Crete. Cross-validation analysis is used to establish the advantages of non-parametric warping for the interpolation of incomplete data. We conclude that wGPR equipped with the proposed data-driven warping provides enhanced flexibility and—at least for the cases studied– improved predictive accuracy for non-Gaussian data. |
| first_indexed | 2024-03-09T13:44:24Z |
| format | Article |
| id | doaj.art-9f2a436da8df41bd8c5e55d9e544b552 |
| institution | Directory Open Access Journal |
| issn | 1099-4300 |
| language | English |
| last_indexed | 2024-03-09T13:44:24Z |
| publishDate | 2022-02-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Entropy |
| spelling | doaj.art-9f2a436da8df41bd8c5e55d9e544b5522023-11-30T21:02:50ZengMDPI AGEntropy1099-43002022-02-0124332110.3390/e24030321Spatial Modeling of Precipitation Based on Data-Driven Warping of Gaussian ProcessesVasiliki D. Agou0Andrew Pavlides1Dionissios T. Hristopulos2School of Mineral Resources Engineering, Technical University of Crete, 73100 Chania, Crete, GreeceSchool of Mineral Resources Engineering, Technical University of Crete, 73100 Chania, Crete, GreeceSchool of Electrical and Computer Engineering, Technical University of Crete, 73100 Chania, Crete, GreeceModeling and forecasting spatiotemporal patterns of precipitation is crucial for managing water resources and mitigating water-related hazards. Globally valid spatiotemporal models of precipitation are not available. This is due to the intermittent nature, non-Gaussian distribution, and complex geographical dependence of precipitation processes. Herein we propose a data-driven model of precipitation amount which employs a novel, data-driven (non-parametric) implementation of warped Gaussian processes. We investigate the proposed warped Gaussian process regression (wGPR) using (i) a synthetic test function contaminated with non-Gaussian noise and (ii) a reanalysis dataset of monthly precipitation from the Mediterranean island of Crete. Cross-validation analysis is used to establish the advantages of non-parametric warping for the interpolation of incomplete data. We conclude that wGPR equipped with the proposed data-driven warping provides enhanced flexibility and—at least for the cases studied– improved predictive accuracy for non-Gaussian data.https://www.mdpi.com/1099-4300/24/3/321non-Gaussian dataskewed distributionsGaussian anamorphosisreanalysis datakrigingwarped Gaussian processes |
| spellingShingle | Vasiliki D. Agou Andrew Pavlides Dionissios T. Hristopulos Spatial Modeling of Precipitation Based on Data-Driven Warping of Gaussian Processes Entropy non-Gaussian data skewed distributions Gaussian anamorphosis reanalysis data kriging warped Gaussian processes |
| title | Spatial Modeling of Precipitation Based on Data-Driven Warping of Gaussian Processes |
| title_full | Spatial Modeling of Precipitation Based on Data-Driven Warping of Gaussian Processes |
| title_fullStr | Spatial Modeling of Precipitation Based on Data-Driven Warping of Gaussian Processes |
| title_full_unstemmed | Spatial Modeling of Precipitation Based on Data-Driven Warping of Gaussian Processes |
| title_short | Spatial Modeling of Precipitation Based on Data-Driven Warping of Gaussian Processes |
| title_sort | spatial modeling of precipitation based on data driven warping of gaussian processes |
| topic | non-Gaussian data skewed distributions Gaussian anamorphosis reanalysis data kriging warped Gaussian processes |
| url | https://www.mdpi.com/1099-4300/24/3/321 |
| work_keys_str_mv | AT vasilikidagou spatialmodelingofprecipitationbasedondatadrivenwarpingofgaussianprocesses AT andrewpavlides spatialmodelingofprecipitationbasedondatadrivenwarpingofgaussianprocesses AT dionissiosthristopulos spatialmodelingofprecipitationbasedondatadrivenwarpingofgaussianprocesses |