The increasing global population inevitably demands for stable food production. As an important food crop, rice plays a major role in maintaining food security. However, irrigated rice fields are increasingly suffered from natural hazard occurrences worldwide, disrupting livelihoods of millions of people and jeopardizing food security. Therefore, there is an urgent need to reduce devastating disaster impacts in irrigated rice fields. In this respect, the concept and practice of disaster risk reduction (DRR) offer insights for reducing damages and losses from natural hazards. As a concept, DRR analyzes and manages causal factors of disaster events, including environmental and socioeconomic processes. In practice, DRR focuses on proactive activities of managing disaster risk instead of solely reacting to disaster impacts. Despite the benefits, surprisingly, few studies incorporate the concept and practice of DRR for investigating or proposing insights to reduce potential disaster impacts in irrigated rice fields. The central question underlying the thesis is “How can disaster impacts in irrigated rice fields be reduced?” This thesis analyzes hazard, vulnerability, and resilience in irrigated rice fields in West Java to answer the question. The area under investigation consists of 4 districts East of Jakarta, containing approximately 37% of the total area of rice fields in West Java. This study uses primary and secondary data, such as stakeholders’ responses and time-series MODIS imageries (MOD09A1). Both quantitative (e.g., remote sensing and statistical analyses) and qualitative approaches (e.g., qualitative content analysis of the interviews) were done. Findings presented in each chapter of this thesis can be used as inputs for designing effective strategies to reduce potential disaster impacts in irrigated rice fields. Information on vulnerability to flooding is essential for estimating potential damages from flood events. Chapter 2 demonstrates that cropping patterns can be used as one of the inputs for deriving physical vulnerability to flooding in irrigated rice fields. Cropping patterns were generated from the spatial distribution of the multiannual Enhanced Vegetation Index, where the timeseries were clustered in areas with similar patterns of crop growth supported by local knowledge on phenology metrics. There is a large spatial and temporal variability in cropping schedules in the area, and at the same moment fields can have freshly planted to fully grown rice. The clusters can be roughly separated in 5 regions from North to South, with cropping patterns related to the timing and availability of irrigation water from the large reservoirs in the South of the area, and landscape position. Combined with vulnerability (damage) curves, cropping patterns can be used to determine vulnerability to flooding. Vulnerability varies in space and time and may shift because of extreme weather variabilities or human decisions. To understand the uncertainty sue to the low resolution, accuracy assessments were performed for the estimated spatial distribution and phenology metrics. For the former, the comparison between MOD09A1 and ALOS PALSAR (2010) and between MOD09A1 and Agricultural Statistics showed coherent results with R2 of 0.81 and 0.93, respectively. For the latter, the estimated RMSEs for SOS, heading stage, and EOS are 9.21 (n=61), 9.29 (n=46), and 9.69 (n=49) days, respectively. Robust flood detection methods are needed for understanding irrigated rice field areas affected by flood events. Previous studies suggested that EVI ≤ 0.1 can be used to detect flood events in irrigated rice fields. However, nonhazardous agronomic inundation needed for rice growth, and hazardous flooding may be present at the same time in irrigated rice fields. Therefore, EVI ≤ 0.1 may not be adequate to detect rice fields with flooding, and the attempt for distinguishing between rice fields with flooding and rice fields with agronomic inundation (RFAI) is therefore necessary. It was found that EVI ≤ 0.1 alone cannot distinguish between flooding and agronomic inundation in irrigated rice fields in the study area. However, when the period of EVI < 0.1 extends beyond 40 days in at the start of the growing season (EVI40) hazardous flood events can be recognized. However, misclassified flood pixels exist partly due to environmental processes, human decisions, and mixed pixels. Using the Start of Season (SOS) for assessing the accuracy of the derived flood map, it is estimated that the scores of Accuracy and F1 for EVI40 are 75.96% and 81.74%, respectively. In other moments of the growing season when the fields have a mature crop, an sudden drop of EVI below 0.1 may be a sign of hazardous flooding. The vulnerability of farmers to natural hazards may partly be explained by unsafe conditions. According to the Pressure and Release (PAR) model, vulnerability may progress from root causes, dynamic pressures to unsafe conditions. Some of the challenges on identifying unsafe conditions are the difference in ‘vocabularies’ among rice agriculture stakeholders (e.g., farmers, water managers, extension officers) and various potential reasons for unsafe conditions. In this respect, disruptions in cropping schedules may be used as a ‘common language’ to understand mechanisms of how unsafe conditions may increase vulnerability. Reasons for disruptions in cropping schedules have been identified, including economic motives, weather variabilities, geographic locations, coping strategies, farmers’ interactions, and gricultural infrastructures. Unsafe conditions in irrigated rice fields in West Java has also been successfully documented, including dangerous locations, unsustainable farming activities, unsuitable coping strategies, fragile infrastructures, and inadequate knowledge and perception of the problems.
|Qualification||Doctor of Philosophy|
|Award date||30 May 2018|
|Place of Publication||Enschede|
|Publication status||Published - 2018|
Sianturi, R. S. (2018). Reducing Potential Disaster Impacts in Irrigated Rice Fields in West Java. Enschede: University of Twente, Faculty of Geo-Information Science and Earth Observation (ITC). https://doi.org/10.3990/1.9789036545549