Dynamic Forecast of Desert Locust Presence Using Machine Learning with a Multivariate Time Lag Sliding Window Technique
Desert locust plagues can easily cause a regional food crisis and thus affect social stability. Preventive control of the disaster highlights the early detection of hopper gregarization before they form devastating swarms. However, the response of hopper band emergence to environmental fluctuation e...
| Main Authors: | , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-02-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-4292/14/3/747 |
| _version_ | 1827658937834405888 |
|---|---|
| author | Ruiqi Sun Wenjiang Huang Yingying Dong Longlong Zhao Biyao Zhang Huiqin Ma Yun Geng Chao Ruan Naichen Xing Xidong Chen Xueling Li |
| author_facet | Ruiqi Sun Wenjiang Huang Yingying Dong Longlong Zhao Biyao Zhang Huiqin Ma Yun Geng Chao Ruan Naichen Xing Xidong Chen Xueling Li |
| author_sort | Ruiqi Sun |
| collection | DOAJ |
| description | Desert locust plagues can easily cause a regional food crisis and thus affect social stability. Preventive control of the disaster highlights the early detection of hopper gregarization before they form devastating swarms. However, the response of hopper band emergence to environmental fluctuation exhibits a time lag. To realize the dynamic forecast of band occurrence with optimal temporal predictors, we proposed an SVM-based model with a temporal sliding window technique by coupling multisource time-series imagery with historical locust ground survey observations from between 2000–2020. The sliding window method was based on a lagging variable importance ranking used to analyze the temporal organization of environmental indicators in band-forming sequences and eventually facilitate the early prediction of band emergence. Statistical results show that hopper bands are more likely to occur within 41–64 days after increased rainfall; soil moisture dynamics increasing by approximately 0.05 m³/m³ then decreasing may enhance the chance of observing bands after 73–80 days. While sparse vegetation areas with NDVI increasing from 0.18 to 0.25 tend to witness bands after 17–40 days. The forecast model combining the optimal time lags of these dynamic indicators with other static indicators allows for a 16-day extended outlook of band presence in Somalia, Ethiopia, and Kenya. Monthly predictions from February to December 2020 display an overall accuracy of 77.46%, with an average ROC-AUC of 0.767 and a mean F-score close to 0.772. The multivariate forecast framework based on the lagging effect can realize the early warning of band presence in different spatiotemporal scenarios, supporting early decisions and response strategies for desert locust preventive management. |
| first_indexed | 2024-03-09T23:11:36Z |
| format | Article |
| id | doaj.art-456f925f30aa49a397da6a7e12b5d7a4 |
| institution | Directory Open Access Journal |
| issn | 2072-4292 |
| language | English |
| last_indexed | 2024-03-09T23:11:36Z |
| publishDate | 2022-02-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj.art-456f925f30aa49a397da6a7e12b5d7a42023-11-23T17:42:50ZengMDPI AGRemote Sensing2072-42922022-02-0114374710.3390/rs14030747Dynamic Forecast of Desert Locust Presence Using Machine Learning with a Multivariate Time Lag Sliding Window TechniqueRuiqi Sun0Wenjiang Huang1Yingying Dong2Longlong Zhao3Biyao Zhang4Huiqin Ma5Yun Geng6Chao Ruan7Naichen Xing8Xidong Chen9Xueling Li10State Key Laboratory of Remote Sensing Science, Aerospace Information Research Institute, Chinese Academy of Sciences, Beijing 100094, ChinaState Key Laboratory of Remote Sensing Science, Aerospace Information Research Institute, Chinese Academy of Sciences, Beijing 100094, ChinaState Key Laboratory of Remote Sensing Science, Aerospace Information Research Institute, Chinese Academy of Sciences, Beijing 100094, ChinaShenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, Shenzhen 518055, ChinaState Key Laboratory of Remote Sensing Science, Aerospace Information Research Institute, Chinese Academy of Sciences, Beijing 100094, ChinaState Key Laboratory of Remote Sensing Science, Aerospace Information Research Institute, Chinese Academy of Sciences, Beijing 100094, ChinaState Key Laboratory of Remote Sensing Science, Aerospace Information Research Institute, Chinese Academy of Sciences, Beijing 100094, ChinaState Key Laboratory of Remote Sensing Science, Aerospace Information Research Institute, Chinese Academy of Sciences, Beijing 100094, ChinaChina Aero Geophysical Survey and Remote Sensing Center for Natural Resources, Beijing 100083, ChinaNorth China University of Water Resources and Electric Power, Zhengzhou 450046, ChinaState Key Laboratory of Remote Sensing Science, Aerospace Information Research Institute, Chinese Academy of Sciences, Beijing 100094, ChinaDesert locust plagues can easily cause a regional food crisis and thus affect social stability. Preventive control of the disaster highlights the early detection of hopper gregarization before they form devastating swarms. However, the response of hopper band emergence to environmental fluctuation exhibits a time lag. To realize the dynamic forecast of band occurrence with optimal temporal predictors, we proposed an SVM-based model with a temporal sliding window technique by coupling multisource time-series imagery with historical locust ground survey observations from between 2000–2020. The sliding window method was based on a lagging variable importance ranking used to analyze the temporal organization of environmental indicators in band-forming sequences and eventually facilitate the early prediction of band emergence. Statistical results show that hopper bands are more likely to occur within 41–64 days after increased rainfall; soil moisture dynamics increasing by approximately 0.05 m³/m³ then decreasing may enhance the chance of observing bands after 73–80 days. While sparse vegetation areas with NDVI increasing from 0.18 to 0.25 tend to witness bands after 17–40 days. The forecast model combining the optimal time lags of these dynamic indicators with other static indicators allows for a 16-day extended outlook of band presence in Somalia, Ethiopia, and Kenya. Monthly predictions from February to December 2020 display an overall accuracy of 77.46%, with an average ROC-AUC of 0.767 and a mean F-score close to 0.772. The multivariate forecast framework based on the lagging effect can realize the early warning of band presence in different spatiotemporal scenarios, supporting early decisions and response strategies for desert locust preventive management.https://www.mdpi.com/2072-4292/14/3/747desert locustenvironmental indicatordynamic forecastmachine learningtime lag |
| spellingShingle | Ruiqi Sun Wenjiang Huang Yingying Dong Longlong Zhao Biyao Zhang Huiqin Ma Yun Geng Chao Ruan Naichen Xing Xidong Chen Xueling Li Dynamic Forecast of Desert Locust Presence Using Machine Learning with a Multivariate Time Lag Sliding Window Technique Remote Sensing desert locust environmental indicator dynamic forecast machine learning time lag |
| title | Dynamic Forecast of Desert Locust Presence Using Machine Learning with a Multivariate Time Lag Sliding Window Technique |
| title_full | Dynamic Forecast of Desert Locust Presence Using Machine Learning with a Multivariate Time Lag Sliding Window Technique |
| title_fullStr | Dynamic Forecast of Desert Locust Presence Using Machine Learning with a Multivariate Time Lag Sliding Window Technique |
| title_full_unstemmed | Dynamic Forecast of Desert Locust Presence Using Machine Learning with a Multivariate Time Lag Sliding Window Technique |
| title_short | Dynamic Forecast of Desert Locust Presence Using Machine Learning with a Multivariate Time Lag Sliding Window Technique |
| title_sort | dynamic forecast of desert locust presence using machine learning with a multivariate time lag sliding window technique |
| topic | desert locust environmental indicator dynamic forecast machine learning time lag |
| url | https://www.mdpi.com/2072-4292/14/3/747 |
| work_keys_str_mv | AT ruiqisun dynamicforecastofdesertlocustpresenceusingmachinelearningwithamultivariatetimelagslidingwindowtechnique AT wenjianghuang dynamicforecastofdesertlocustpresenceusingmachinelearningwithamultivariatetimelagslidingwindowtechnique AT yingyingdong dynamicforecastofdesertlocustpresenceusingmachinelearningwithamultivariatetimelagslidingwindowtechnique AT longlongzhao dynamicforecastofdesertlocustpresenceusingmachinelearningwithamultivariatetimelagslidingwindowtechnique AT biyaozhang dynamicforecastofdesertlocustpresenceusingmachinelearningwithamultivariatetimelagslidingwindowtechnique AT huiqinma dynamicforecastofdesertlocustpresenceusingmachinelearningwithamultivariatetimelagslidingwindowtechnique AT yungeng dynamicforecastofdesertlocustpresenceusingmachinelearningwithamultivariatetimelagslidingwindowtechnique AT chaoruan dynamicforecastofdesertlocustpresenceusingmachinelearningwithamultivariatetimelagslidingwindowtechnique AT naichenxing dynamicforecastofdesertlocustpresenceusingmachinelearningwithamultivariatetimelagslidingwindowtechnique AT xidongchen dynamicforecastofdesertlocustpresenceusingmachinelearningwithamultivariatetimelagslidingwindowtechnique AT xuelingli dynamicforecastofdesertlocustpresenceusingmachinelearningwithamultivariatetimelagslidingwindowtechnique |