Drought risk analysis and rainfall index insurance model development
Drought has been identified as the main threat to agriculture production and people’s lives, due to its long-term social and economic impact. The frequency analysis of drought risk plays an important role in helping insurers to (a) identify the spatial distribution of risk in the insured area, (b) m...
Saved in:
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Theses and Dissertations |
| Language: | English |
| Published: |
2015
|
| Subjects: | |
| Online Access: | https://hdl.handle.net/10356/65574 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Nanyang Technological University |
| Language: | English |
| Summary: | Drought has been identified as the main threat to agriculture production and people’s lives, due to its long-term social and economic impact. The frequency analysis of drought risk plays an important role in helping insurers to (a) identify the spatial distribution of risk in the insured area, (b) make decisions about risk pooling, and (c) calculate the potential extreme losses. This research project is the first in the field of Standardized Precipitation Index (SPI)–based drought-risk studies to use diffusion kernel density estimation (DKDE) to estimate the bivariate probability density functions (PDFs) and the joint return period (RP). Historical daily rainfall data collected from 19 weather stations in China’s Shandong province was used to assess the DKDE for the drought-risk frequency analysis in this thesis. The results show that the DKDE method is capable of producing an index that can consider multiple factors with higher accuracy, and of eliminating the unwanted probability shifts that occur in the existing estimation method, at a lower computation cost. The utilization of the DKDE function for drought-risk analysis also provides a reference for identifying regional agricultural drought, and offers important technological support for drought-risk management. Agriculture insurance plays an important role in compensating farmers for revenue loss due to adverse weather events. Rainfall index insurance, based on the assumption that crop productivity and income from farmers are highly correlated with precipitation in the crop-growth phases, has attracted the attention of researchers and institutions for its relatively lower transaction cost, faster loss adjustment, and faster payout. A rainfall index insurance model was developed for this research based on the statistical analysis of the relationship between rainfall per corn phenological growth phase and yield reduction. This model distinguishes itself from the existing approaches by independently considering the characteristics of each growth phase in different insured areas for correlation development. The Probable Maximum Loss (PML) caused by drought was estimated through calculating the drought severity extreme return level at r-return period using DKDE. The cumulative rainfall (CR) index insurance model was established for five counties of Shandong province with historical daily rainfall data (1981–2011) and corn-yield data (1985–2011). The results of premium and premium rate pricing suggest that rainfall index insurance is a viable product to complement the existing indemnity-based, government-supported corn insurance program in Shandong province. Furthermore, rainfall indices could be a developed into a possible product to insure farmers against drought in regions where no insurance coverage is currently available, due to high historical drought-caused yield losses. |
|---|