Evaluation of machine learning approaches for large-scale agricultural drought forecasts to improve monitoring and preparedness in Brazil
<p>Drought events have increased in frequency and severity in recent years and result in significant economic losses. Although the Brazilian semi-arid Northeast has been historically associated with the impacts of drought, drought is of national concern. From 2011–2019, drought events w...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Copernicus Publications
2025-04-01
|
| Series: | Natural Hazards and Earth System Sciences |
| Online Access: | https://nhess.copernicus.org/articles/25/1521/2025/nhess-25-1521-2025.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849712285693509632 |
|---|---|
| author | J. W. Gallear M. Valadares Galdos M. Zeri A. Hartley |
| author_facet | J. W. Gallear M. Valadares Galdos M. Zeri A. Hartley |
| author_sort | J. W. Gallear |
| collection | DOAJ |
| description | <p>Drought events have increased in frequency and severity in recent years and result in significant economic losses. Although the Brazilian semi-arid Northeast has been historically associated with the impacts of drought, drought is of national concern. From 2011–2019, drought events were recorded in all Brazilian territories. Droughts can have major consequences for agricultural production, which is of particular concern given the importance of soybeans for socio-economic development. Due to its regional heterogeneity, it is important to develop accurate drought forecast and assessment tools for Brazil. We explore machine learning as a method to forecast the vegetation health index (VHI), for large-scale monthly drought monitoring across agricultural land in Brazil. Furthermore, we also determine spatio-temporal drivers of the VHI across the wide variation in climates and evaluate machine learning performance for El Niño–Southern Oscillation variation and forecasting of the onset of drought stress. We show that machine learning methods such as gradient boosting methods are able to more easily forecast vegetation health in north and northeast Brazil than in south Brazil, and they perform better during La Niña events than El Niño events. Drought stress which reduces the VHI below the commonly used 40 % threshold can be forecast across Brazil with similar model performance. The standardized precipitation evapotranspiration index is shown to be a useful indicator of drought stress, with 3-month accumulation periods preferred over 1- and 2-month periods. Results aim to inform future developments in operational drought monitoring at the National Centre for Monitoring and Early Warning of Natural Disasters in Brazil (CEMADEN). Future work should build upon methods discussed here to improve drought forecasts for agricultural drought response and disaster risk reduction.</p> |
| format | Article |
| id | doaj-art-701e99fb2c6b4f679ab8d5cee3edccc9 |
| institution | DOAJ |
| issn | 1561-8633 1684-9981 |
| language | English |
| publishDate | 2025-04-01 |
| publisher | Copernicus Publications |
| record_format | Article |
| series | Natural Hazards and Earth System Sciences |
| spelling | doaj-art-701e99fb2c6b4f679ab8d5cee3edccc92025-08-20T03:14:19ZengCopernicus PublicationsNatural Hazards and Earth System Sciences1561-86331684-99812025-04-01251521154110.5194/nhess-25-1521-2025Evaluation of machine learning approaches for large-scale agricultural drought forecasts to improve monitoring and preparedness in BrazilJ. W. Gallear0M. Valadares Galdos1M. Zeri2A. Hartley3Rothamsted Research, West Common, Harpenden, UKRothamsted Research, West Common, Harpenden, UKNational Centre for Monitoring and Early Warning of Natural Disasters (CEMADEN), São José dos Campos, BrazilMet Office Hadley Centre, FitzRoy Road, Exeter, UK<p>Drought events have increased in frequency and severity in recent years and result in significant economic losses. Although the Brazilian semi-arid Northeast has been historically associated with the impacts of drought, drought is of national concern. From 2011–2019, drought events were recorded in all Brazilian territories. Droughts can have major consequences for agricultural production, which is of particular concern given the importance of soybeans for socio-economic development. Due to its regional heterogeneity, it is important to develop accurate drought forecast and assessment tools for Brazil. We explore machine learning as a method to forecast the vegetation health index (VHI), for large-scale monthly drought monitoring across agricultural land in Brazil. Furthermore, we also determine spatio-temporal drivers of the VHI across the wide variation in climates and evaluate machine learning performance for El Niño–Southern Oscillation variation and forecasting of the onset of drought stress. We show that machine learning methods such as gradient boosting methods are able to more easily forecast vegetation health in north and northeast Brazil than in south Brazil, and they perform better during La Niña events than El Niño events. Drought stress which reduces the VHI below the commonly used 40 % threshold can be forecast across Brazil with similar model performance. The standardized precipitation evapotranspiration index is shown to be a useful indicator of drought stress, with 3-month accumulation periods preferred over 1- and 2-month periods. Results aim to inform future developments in operational drought monitoring at the National Centre for Monitoring and Early Warning of Natural Disasters in Brazil (CEMADEN). Future work should build upon methods discussed here to improve drought forecasts for agricultural drought response and disaster risk reduction.</p>https://nhess.copernicus.org/articles/25/1521/2025/nhess-25-1521-2025.pdf |
| spellingShingle | J. W. Gallear M. Valadares Galdos M. Zeri A. Hartley Evaluation of machine learning approaches for large-scale agricultural drought forecasts to improve monitoring and preparedness in Brazil Natural Hazards and Earth System Sciences |
| title | Evaluation of machine learning approaches for large-scale agricultural drought forecasts to improve monitoring and preparedness in Brazil |
| title_full | Evaluation of machine learning approaches for large-scale agricultural drought forecasts to improve monitoring and preparedness in Brazil |
| title_fullStr | Evaluation of machine learning approaches for large-scale agricultural drought forecasts to improve monitoring and preparedness in Brazil |
| title_full_unstemmed | Evaluation of machine learning approaches for large-scale agricultural drought forecasts to improve monitoring and preparedness in Brazil |
| title_short | Evaluation of machine learning approaches for large-scale agricultural drought forecasts to improve monitoring and preparedness in Brazil |
| title_sort | evaluation of machine learning approaches for large scale agricultural drought forecasts to improve monitoring and preparedness in brazil |
| url | https://nhess.copernicus.org/articles/25/1521/2025/nhess-25-1521-2025.pdf |
| work_keys_str_mv | AT jwgallear evaluationofmachinelearningapproachesforlargescaleagriculturaldroughtforecaststoimprovemonitoringandpreparednessinbrazil AT mvaladaresgaldos evaluationofmachinelearningapproachesforlargescaleagriculturaldroughtforecaststoimprovemonitoringandpreparednessinbrazil AT mzeri evaluationofmachinelearningapproachesforlargescaleagriculturaldroughtforecaststoimprovemonitoringandpreparednessinbrazil AT ahartley evaluationofmachinelearningapproachesforlargescaleagriculturaldroughtforecaststoimprovemonitoringandpreparednessinbrazil |