RSS | Register/註冊 | Log in/登入
Site search:
Home>FFTC Document Database>Extension Bulletins>DEVELOPING DISASTER EARLY WARNING SYSTEM AND ADAPTATION STRATEGIES FOR CROP PRODUCTION IN TAIWAN
facebook分享

Ming-Hwi Yao, Tzay-Ming Leou1, Yung-Heng Hsu2, Chi-ling Chen3, Chun-Tang Lu4

Agricultural Engineering Division,

Taiwan Agricultural Research Institute (TARI), Wufeng, Taichung, Taiwan

1Central Weather Bureau, Taipei, Taiwan

2National Science and Technology Center for Disaster Reduction, Taipei, Taiwan

3Agricultural Chemistry Division,

Taiwan Agricultural Research Institute (TARI),

Wufeng, Taichung, Taiwan

4Crop Science Division,

Taiwan Agricultural Research Institute (TARI),

Wufeng, Taichung, Taiwan

 

ABSTRACT

Agriculture is an industry that strongly depends on the environment. Recently, it has been especially affected by an increase in the frequency of extreme disasters caused by climate change. Such disasters have an impact on agricultural lands and the stable supply of agricultural products. Therefore, the Taiwanese government implemented the agricultural meteorological disasters adaptation strategy research program to strongly promote disaster prevention information and techniques and to provide refined forecasting as well as an early warning system to establish preparedness and recovery in agricultural disasters. Moreover, this measure could convert passive mitigation into active action through disseminating information to farmers. This study illustrated the current situation of agricultural disasters in Taiwan in terms of three aspects: crop losses caused by agricultural disasters, information and communication technology for agricultural disaster prevention, and research and development in disaster prevention techniques.

Keywords: disaster, crop, early warning system, information and communication technology

INTRODUCTION

Climate change leads to frequent extreme weather conditions, which has an immediate impact on agricultural production. Scientists have analyzed the effects of a rise in temperature since 1980 and have suggested that this rise is closely related to heavy rain events; 26% of extreme rainfall events are clearly related to climate change. According to the global climate risk index released by Germany’s Germanwatch organization in 2016, the main cause of disasters in the 10 countries with high climate risk is extreme rain (including typhoon). Taiwan ranked 39 out of 183 countries with moderate climate risk. In the past few decades, the number of disasters has increased steadily, and the damage has become more severe, which has led to the destruction of infrastructure and the natural environment. However, the purpose of disaster management is to reduce or avoid the potential threat of disaster, provide immediate and accurate assistance to victims, and improve disaster prevention through message coordination and transfer. Disaster management involves predisaster risk reduction and postdisaster recovery (Seneviratne et al., 2010). Predisaster risk reduction can be achieved using an early warning system and a standard operating procedure; an early warning system is a cost-effective and high-efficiency strategy to save more property and lives (Rogers and Tsirkunov, 2010). A complete early warning system must involve the following: 1. continuous monitoring and real-time updates; a forecaster must understand potential threats of disaster; 2. high accuracy of forecast data; 3. appropriate announcement methods to ensure that information is promptly disseminated; and 4. actual and high coordination of the receiver to maximize the effectiveness of the early warning system (Tembo et al., 2014).

Frost damage is one of the main causes of serious crop losses in the United States. A complete early warning system could help minimize the potential damage caused by frost and frostbite. Disaster forecast depends on several factors, including atmospheric temperature, dew point temperature, and wind speed. In addition, years of expert experience is required to thoroughly establish the system. Georgia’s extreme-weather neural-network informed expert model incorporates the knowledge of expert agrometeorologists and historical weather data on air temperature, dew point temperature, and wind speed into the system to provide warnings of frost levels for blueberry and peach farming. This tool is available through a web-based interface (Chevalier et al., 2012). Drought and waterlogging are the main disasters that restrict agricultural production in China (Li et al., 2000). Northeastern China is one of the major contributors to the Maize Belt in the world. Although the climatic conditions of this region are favorable to maize growth, disasters, especially drought, occur frequently and severely affect crop production. For example, in Liaoning Province, drought accounts for 60% of all types of disasters. Moreover, the frequency of droughts often increases with global warming. A period of drought continued from spring until autumn in 2009, which caused a reduction in maize production (Zhang, 2004; Wilhite et al., 2007). Zhang et al. (2014) used the geographical information system (GIS) to downscale Liaoning Province into small grids with higher resolution. Based on the daily meteorological data and maize yield data from 1997 to 2005, the probability of drought was calculated using logistic regression. Then, by considering precipitation, wind speed, and temperature as the threshold, the critical value of drought level was defined, and the early warning system was established, which provided a reference to government agencies to implement emergency strategies to reduce crop losses.

Agricultural production in Taiwan is often hindered by meteorological disasters, with annual losses of approximately 1–27 billion NTD. This leads to no income to farmers and affects the stable supply of agricultural products, causing consumer complaints. Common agrometeorological disasters include typhoons, heavy rain, cold damage, and drought. Therefore, the Taiwanese government implemented the agricultural meteorological disasters adaptation strategy research program to strongly promote disaster prevention information and techniques and to provide refined forecasting and an early warning system to establish preparedness and recovery in agricultural disasters. Moreover, this measure could convert passive mitigation into active action through disseminating information to farmers. This study illustrated the current situation of agricultural disasters in Taiwan from three aspects: crop losses caused by agricultural disasters, information and communication technology (ICT) for agricultural disaster prevention, and research and development in disaster prevention techniques.

Taiwan is an island with a rugged mountain landscape. It is often hit by many meteorological disasters, especially heavy rains and typhoons. The annual crop losses caused by agrometeorological disasters are approximately $NT5 billion in Taiwan (Fig. 1). Fig. 2 illustrates the statistics of crop losses of the past 30 years from 1985 to 2014. Typhoons are the most serious threats to crop production accounting for approximately 63.6% of the damage; rain damage includes that from spring rain, the East Asian rainy season, and convectional rain, accounting for 11.7%; the first rice crop, fruit trees, and tea are affected by cold damage, accounting for approximately 8.2%; hailstones, drought, high temperature damage, and so called foehn winds account for the remaining proportion of damage. An analysis revealed that the crops most affected by losses due to damage are fruits, accounting for 47.9%; vegetables, accounting for 28.0%; and rice, accounting for approximately 10.8% (Fig. 3). Damage to the crop is subject to the crop’s growth stage, planting environment, and sensitivity to the disaster. According to crop loss statistics, the farming period and region greatly influence severe crop losses. Moreover, a comparison between the damages caused by the cold wave in February 2016 and those caused by the cold waves in 2005 and 2008 revealed that the affected crops and regions were similar, indicating that damages caused by disasters have similar patterns. Therefore, determining the hot zones for different crops could facilitate predisaster prevention and postdisaster recovery.

Fig. 1. Historical crop losses caused by meteorological disasters in Taiwan.

Fig. 2. Crop losses caused by various disasters in Taiwan (values and percentages during 1985–2014).

Fig. 3. Crop losses caused by disasters in Taiwan (values and percentages during 1985–2014).

 

APPLICATION OF ICT IN AGROMETEOROLOGICAL DISASTER INFORMATION

The application of ICT to agrometeorological disaster information is critical for disaster prevention, but the information source must be analyzed and integrated to meet the disaster prevention use. This section describes the application of ICT to the disaster prevention system at different times.

Predisaster meteorological information

The information on weather forecast and disaster alerts depends on real-time observation data. Taiwan established an agrometeorological website, which includes data from 17 agrometeorological stations (data delivered through the network in time) and 11 agrometeorological stations (paper report sent back to Central Weather Bureau (CWB) every 10 days). However, the number of agrometeorological stations is still insufficient; more stations must be established in agricultural regions to enhance disaster prevention. Therefore, the research program implemented by the Taiwanese government upgraded the former agrometeorological stations and established 100 new stations as of 2017. This will enhance the forecasting capabilities for agricultural disasters; moreover, farmers can make decisions by referring to the website (http://agr.cwb.gov.tw) (Fig. 4). Currently, CWB provides weather forecasting at the township spatial scale; however, farmers require more refined information at the spatial and temporal scales. To achieve this, the Council of Agriculture and CWB collaboratively planned important crop cultivation areas to provide weather forecasting at a scale of 2.5 × 2.5 km2 and developed a long-term forecasting system, hoping to meet farmers’ demand for meteorological information.

Fig. 4.  Home page of the agrometeorological monitoring system (http://agr.cwb.gov.tw).

After analyzing the vulnerability and hot zones of different crops, the disaster information and resources can be used to properly allocate the disaster-prone areas. In this research program, a monthly probability map of agrometeorological disasters was drawn based on the historical weather data and a probability map of crop losses was drawn based on historical crop losses data with the current cultivation area and land-use map. These maps marked the hot zones of the economically important crops, illustrated real-time disaster alerting regions, and provided prevention measures to reduce losses. Simultaneously, the production fragility of agricultural areas and economic losses caused by the disasters were analyzed, and suitable cropping areas with a view of disaster prevention were established.

The sensitivity to weather differs between crops. In addition to the genetic characteristics of crops, the perception of weather conditions at different growth stages influences the sensitivity of crops to weather. For example, rice is highly tolerant to high temperatures (>35 °C) during the vegetative growth stage, but high temperatures may cause sterility or low quality during the flowering or grain-filling stages. Moreover, elucidating the disaster probability at each growth stage of crops and critical thresholds through simulation experiments, literature reviews, and farmer interviews could provide a reference for disaster warnings. In addition, farmers could easily obtain information on distinct “disaster prevention calendars” for economically important crops, including the monthly growth stage, impending disasters, critical thresholds, and prevention measures.

Combining the database of critical thresholds of crops and refined forecasts from the CWB to develop the “crop disaster early warning system” (http://disaster.tari.gov.tw) (Fig. 5) could help automatically determine the severity of agrometeorological disasters. In addition, the actual status after disaster can be used to revise the system to determine indicators set as a routine mechanism, providing the mobile application facilitate farmer receive the disaster information easier. In addition to disaster warning and immediate notification, the system also provides the latest activity information; real-time observation data; weather forecast of cultivation areas; disaster information; historical data such as Taiwan’s agricultural climate patterns, disaster rates, and maps of common disaster hot zones; disaster prevention calendar; 24 solar terms; and achievements of the agricultural and forestry disaster prevention program. Rich agricultural meteorological and disaster prevention information will equip farmers to reduce crop losses caused by disasters. Moreover, we developed an application system to facilitate farmers to obtain crop disaster information through smartphones (Fig. 6).

Fig. 5. Home page of the crop disaster early warning system (http://disaster.tari.gov.tw).

Fig. 6. Images of the application and the QR code from the crop disaster early warning system.

Meteorological information during disasters

The National Science and Technology Center for Disaster Reduction developed an “agricultural information service platform for disaster” (http://eocdss.ncdr.nat.gov.tw/web/ot/coa) (Fig. 7) to provide information on the current crop cultivation situation in disaster-prone areas, road situation of agricultural production, and the standard operating procedure of disaster prevention. When the probability of the disaster reaches a certain extent, the agricultural research and extension station or farmers’ association or agricultural production and marketing groups release the disaster information and prevention measures through the Internet, messages, and media; moreover, government agencies could grasp the agricultural disaster information immediately through the weather and hydrological monitoring map data in government disaster response meetings.

Fig. 7. Home page of the agricultural information service platform for disaster (http://eocdss.ncdr.nat.gov.tw/web/ot/coa).

 

Postdisaster meteorological information

The primary tasks after a disaster are provision of emergency allowances and rehabilitation; however, determination of crop losses is controversial. The introduction of unmanned aerial vehicles (UAVs) for determining the current land-use situation, disaster areas, and crop losses facilitates establishment of the image database before and after the disaster with high efficiency. Moreover, it helps to address the controversy and more accurately estimate crop losses. However, it also develops high-precision crop image recognition techniques. To establish the standard operating procedure and basic information on cultivation and disaster areas, UAVs can be employed to obtain high-quality images before and after disasters by using the global positioning system, followed by an analysis of the images by using cadastral data and the GIS. In this study, UAVs were applied to capture instant images for improving the disaster investigation and rescue processes. The study areas were the paddy rice fields in central Taiwan. These paddy fields have been damaged by heavy rains during the monsoon season in June 2016. UAV images provide high ground resolution (3.5 cm) with 3D point clouds to achieve image discrimination and develop a digital surface model (DSM) to determine rice lodging. First, supervised classification of the images was performed using the maximum likelihood method to delineate the rice lodging area. Second, 3D point clouds generated using Pix4D Mapper were used to develop the DSM for classifying the rice lodging levels. We obtained a rice lodging discrimination accuracy of 85% through supervised image classification and a lodging level classification accuracy of 87% by using the DSM. Therefore, UAVs not only provide instant images of agricultural damage after a meteorological disaster but also image discriminations on rice lodging with acceptable accuracy (>85%). In the future, UAV and image discrimination technologies can be applied to different crop fields. The results of image discrimination were overlapped with the administrative boundaries of rice paddy to establish a GIS-based assistance system for agricultural damage discrimination. Therefore, the time and labor required for damage detection and monitoring was greatly reduced. Figure 8 presents the flowchart of agricultural disaster image interpretation by using UAVs and the image analysis technique in this study.

Fig. 8. Flowchart of the interpretation of agricultural disaster images obtained using UAVs and the image analysis technique.

ICT is widely used in meteorological data collection and other functions in Taiwan. For example, the physiological responses of crops to various stresses were used to establish an early warning system, and information was integrated to provide disaster prevention measures. However, too much information and complicated operating procedures may cause users to hesitate to use these products. Therefore, simplified operating procedures such as mobile applications should be used to attract users.

DISASTER PREVENTION AND AVOIDANCE TECHNIQUES

Techniques for preventing and avoiding disasters include selecting economically important crops as targets, reducing the risk of crop losses through the use of facilities, adjusting planting time, and assessing suitability. For crops that require large planting areas and yield a low production value per unit area, disaster prevention can be promoted through crop breeding and fertilizer cultivation management. In some regions, specific crops are prone to disasters; therefore, farmers can be recommended to grow other crops to achieve disaster avoidance. Different disaster prevention techniques or strategies exist corresponding to different growth stages. The research program implemented by the Taiwanese government adopted prevention techniques for disasters that economically important crops might encounter, with emphasis on promotion of developed techniques. For example, southeastern Taiwan is prone to foehn wind phenomenon, characterized by high temperature and low humidity, which cause the flowers or immature fruits to become burnt. This phenomenon often occurs suddenly; thus, notifications to farmers are often sent too late. In fact, farmers usually employ a simple foehn wind sensor device in the sprinkler pipes in their orchards used for fertilizing or watering purposes. When foehn wind occurs, the sensor device detects it and automatically switches on the sprinkler system, which cools the temperature and elevates the humidity level in order to avoid the damage caused by foehn wind. Other techniques such as the use of a windbreak net to reduce damage caused by the strong wind, the use of a spoiler fan to reduce frost damage to tea trees, and field irrigation to reduce the damage caused by cold waves involve low costs, are easy to operate, and could slightly reduce damage.

CONCLUSION

Although disasters have a great impact on crop production, the current strategies for disaster prediction are not sufficiently effective. The accuracy of disaster forecasting cannot be improved in a short time; therefore, promoting disaster information and prevention measures to reduce losses due to damage is the most appropriate strategy. Although some disasters (e.g., hailstones) are difficult to forecast and prevent, for others (e.g., typhoons), there is a critical a prevention time from the release of the forecast to the typhoon landing, during which standard operating procedures for disaster prevention can be conducted. Predisaster prevention is the most efficient and cost-effective disaster adaptation strategy. Early warning systems and corresponding disaster prevention mechanisms established by the agricultural meteorological disasters adaptation strategy research program should be used to implement disaster prevention and prevention measures in conjunction with agricultural producers through education training and workshops. In addition, different crops in Taiwan have different disaster sensitivities. Taiwan has established a disaster prevention calendar of economically important crops, which can be used as reference for crop disaster warning, crop loss assessment, and agricultural insurance management. Additionally, it can also be used as a knowledge base for climate change impact assessment. In the future, this disaster prevention technique can be combined with a system included information on crops, soil, pests and diseases, and cultivation techniques to effectively integrate agricultural production and meteorological resources, accurately reflect the variation in climate at spatial and temporal scales, and implement cultivation management with high efficiency to reduce the impact of disasters on agricultural production.

REFERENCES

Chevalier, R.F., G. Hoogenboom, R.W. McClendon and J.O. Paz. 2012. A web based fuzzy expert system for frost warnings in horticultural crops. Environmental Modeling and Software 35: 84-91.

Li, K.R., Y.F. Chen and C.Y. Huang. 2000. The impacts of drought in China: recent experiences. Drought: A Global Assessment, 331–348.

Rogers, D. and V. Tsirkunov. 2010. The costs and benefits of early warning systems. In: Global assessment report on disaster risk reduction. Published by United Nations Office for Disaster Risk Reduction, Geneva, Switzerland.

Seneviratne, K., D. Baldry and C. Pathirage. 2010. Disaster knowledge factors in managing disasters successfully. International Journal of Strategic Property Management 14: 376-390.

Tembo, G., B. Chimai, N. Tembo and M. Ndiyoi. 2014. Observations on Zambia’s Crop Monitoring and Early Warning Systems. Journal of Agricultural Science 6: 99-107.

Wilhite, D.A., M. Svoboda and M.J. Hayes. 2007. Understanding the complex impacts of drought: a key to enhancing drought mitigation and preparedness. Water Resources Management 21: 763-774.

Zhang J.Q. 2004. Risk assessment of drought disaster in the maize growing region of Songliao plain, China. Agriculture, Ecosystems and Environment 102: 133-153.

Zhang, Q., J. Zhang, C. Wang, L. Cui and D. Yan. 2014. Risk early warning of maize drought disaster in Northwestern Liaoning Province, China. Natural Hazards 72: 701-710.

 

 

 

 


Download the PDF. of this document (80), 896,439 bytes (875 KB).