Figure 3.10. Bangladesh: Monthly Average Maximum and Minimum Temperatures

Source: Authors.

65. 
    Although temperatures are generally homogenous throughout the Bangladeshi territory, lower temperatures are recorded in the northeastern districts while the highest temperatures are recorded in the northwest region of the country (figure 3.11).
    

Source: Authors

66. 
    A weather-indicator-based³⁷ production-risk analysis enables the isolation of simple climate variables (e.g., precipitation, maximum or minimum temperatures, or windspeed) among all the abiotic environmental parameters that influence yields in order to sudy each variable’s positive or negative impact. The selected indicators for this study were designed to capture the potential negative impact of weather-driven stresses on the studied rice varieties that ultimately lead to yield loss.
    
67. 
    As a preliminary weather-risk assessment on rice production, a national-scale rainfall indicator-based analysis was carried out in order to obtain a preliminary overview of the indicators’ ability to the capture weather-driven production risk on Aman HYV rice across Bangladesh. The analysis was conducted based on data from selected Districts from all regions of Bangladesh. All rainfall-based indicators were based on year-wise monthly cumulative precipitation data covering the whole Aman rice growing season (i.e., April to October) in each of the studied districts. Three indicators were built in order to measure (i) cumulative, (ii) deficit, and (iii) excess rainfall. Deficit and excess rainfalls were calculated from the meteorological data in each district based on the deviation from the mean average monthly rainfall. Subsequently, the obtained year-wise indicator data sets were statisitcally analyzed against Aman HYV District-level aggregated yield records in order to detect vulnerabilities of the yield to deficit and/or excess as well as to situate them in time. More detail concerning the structure of the rainfall and other weather-based indicators that were used in the present study can be found in annex 8.
    
68. 
    Map 3.8 presents the results of this analysis and shows its overall consistency with the multicriteria drought risk evaluation and mapping conducted by BARC. Generally in accordance with the BARC drought risk assessment, the present rainfall indicator-based risk assessment found four main categories of rainfall-linked production risk for HYV Aman rice production: (i) acute water deficit vulnerability, (ii) acute excess rainfall vulnerability, (iii) “mild” deficit precipitation sensitivity, and (iv) “mild” excess precipitation sensitivity. It is noteworthy that in three Districts (Dhaka, Comilla, and Chittagong) this preliminary indicator-based risk assessment proved inconclusive, reflecting the complexity of the intertwined nature of environmental paramterers driving yield variation and the limit of indicator-based analysis.