Climate risk in agriculture represents the probability of a defined hydro-meteorological hazard affecting the livelihood of farmers. Risk refers to a probability that can be estimated from prior information, while uncertainty applies to situations in which probability cannot be estimated. Both risks and uncertainties contribute to choice of appropriate management practices by the decision-makers in agriculture. Farmers to some extent understand the risks and uncertainties of climate at their location and optimize the management practices based on years of experience. However, growing demand, changing climatic conditions, intensification and spread of agriculture to marginal production environments warrants improved climate risk management and decision support systems to enable appropriate choice of practices and strategies to match the current and future climate risks.

Bangladesh with a relatively large share of population dependent on agriculture is more prone to the adverse effects of climatic variability. Historical trends and a number of model based projections point to a noticeable increase in surface temperature and erratic rainfall in Bangladesh. This along with the possible changes in spatial and temporal patterns in rainfall poses challenges to sustainable agricultural production. Keeping the need to make Bangladesh agriculture more resilient to changing and increasingly variable climate. In order to develop and target appropriate adaptation measures, it is important to identify regions that are more affected by climatic variability. This identification process involves assessment of vulnerability of different upazilas in Bangladesh. This proposed risk mapping is presented the analysis of vulnerability of agriculture to climate variability and risks at the upazila level considering the fact that most of the development planning and programme implementation is done at upazila level in Bangladesh. The output of the analysis has been presented in the form of maps representing various dimensions of vulnerability. This approach is useful in identifying the upazilas that are vulnerable to climate so that necessary management practices and adoption of technologies could be targeted. It is also useful in identifying sources of vulnerability that are critical to developing appropriate adaptation measures in terms of technologies, investments and polices. With the above background, this study has been conducted with the aim to identify the natural and climate induced vulnerability of the upazilas level to determine the highest vulnerable areas and provide the intervention benefits of the formers against various degrees of risk and vulnerability. Presented below the major conclusions based on the study:

1. This study, shows that most of the cyclone hit around Chottagram-Feni coast and these coastal areas was observed as at very high risk. On the other hand, only high risk shows in Barguna-Patuakhali-Noakhali-Bhola coast, whereas Khulna coast areas show low risk area, and other areas show very low to no risk. Upazilas affected by the cyclone at different intensities in the above-areas.
2. Haor region is found very high to high risk area for flash flood. On the other hand, only high risk area is noticed over the upazilas under Madaripur, Shariatpur, Rajbari, Faridpur, Manikganj, Jamalpur and Gaibandha zilas. The rest of the zilas are comparatively less vulnerable.

1. This study, shows that most of the cyclone hit around Chottagram-Feni coast and these coastal areas was observed as at very high risk. On the other hand, only high risk shows in Barguna-Patuakhali-Noakhali-Bhola coast, whereas Khulna coast areas show low risk area, and other areas show very low to no risk. Upazilas affected by the cyclone at different intensities in the above-areas.
2. Haor region is found very high to high risk area for flash flood. On the other hand, only high risk area is noticed over the upazilas under Madaripur, Shariatpur, Rajbari, Faridpur, Manikganj, Jamalpur and Gaibandha zilas. The rest of the zilas are comparatively less vulnerable.
3. The study observes that very high vulnerable area for cold wave in the upazilas under Rajshahi-Dinajpur regions. The upazilas under Chuadanga, Meherpur, Kushtia, Natore, Bogura, Guaibandha, Rangpur, Kurigram, Nilphamri, Lalmonirhat and Moulvibazar zilas are high vulnerable area. The rest locations (upazilas) is found to be low to moderate vulnerable area for cold wave stress.
4. High to very high heat wave vulnerable area is observed in the upazilas in the western part; most of the upazilas are found to be low to moderate heat wave vulnerable area in Bangladesh except the southern parts costal area; the latter region, is classed as no heat waves vulnerable area.
5. High to very high salinity intrusion is observed over almost all upazilas under Satkhira, Khulna, Barguna, Pirojpur, Jhalokati, Patuakhali, Bhola Noakhali and Cox’s Bazar zilas and low to moderate over the upazilas under Narial, Gopalganj, Madaripur, Shariatpur, Chanpur, Lakshimpur, Feni and Chittagong zilas.
6. The produced vulnerable map with respect to storm surge inundation, the designates very high category upazilas in the mid-southern and southeast part (Patuakhali, Bhola, Noakhali, Chattogram zilas) and high category upazilas in the Satkhira, Khulna, Bagerhat, Chandpur, Lakshimipur, Barishal, Bhola, Narsingdi, Cox’s Bazar, Shariatpur, zilas; the rest coastal locations (upazilas) are classed as low to moderate vulnerable.
7. With respect to the amount of 1-day maximum rainfall, very high risk area is categorized for some upazilas in the southeastern part of Chattogram and Bhola zilas, high risk area is found to be in some upazilas of Mymensingh, Chattogram. Noakhali, Sylhet and Dinajpur zilas, and low to moderate risk area in other locations; this categorization has been based on observed rainfall (BMD data). Using gridded Aphrodite rainfall data, high to very high risk areas have been designated for some upazilas of Sylhet, Sunamganj and Cox’s Bazar zilas, whereas, low to moderate risk areas is demarked over the upazilas of southern and northeastern, Bangladesh and the rest of the upazilas (notably western part) classed as low risk area. With CRU rainfall data, upazilas under Noakhali, Bhola, Sunamganj and Sylhet zilas belonged to very high to high risk areas, whereas, study locations of southeast to central part is designated to be low to moderate risk area. The risk categorization using CRU rainfall was similar to that using the Aphrodite gridded rainfall data-source.
8. Vulnerability Index (SPI) based on monsoon drought analysis using observed gridded Aphrodite rainfall, indicates the prevalence of severe to extremely dry conditions in most of the upazilas over southern and western part of Bangladesh; some upazilas of northwest and some isolated upazilas of eastern part showed mild wet conditions, whereas extreme wet conditions were seen in some upazilas of Sylhet region. Furthermore, mild dry conditions prevailed in some pocket upazilas of northern part of Bangladesh.  The gridded Aphrodite rainfall based SPI was very similar to observed rainfall.

Projections

1. Vulnerability Index (SPI) based on monsoon seasonal drought frequency is likely to decrease over northwest upazilas of Bangladesh under RCP4.5. Mild to moderate drought conditions would prevail over northwest upazilas of Bangladesh. On the other hand, moderate to extremely wet conditions are expected over southern upazilas and mild wet conditions may prevail over narrow part of central upazilas (along Meherpur-Rajshahi to Sylhet-Moulvibazar) in 2030. But in 2050 in southern regional upazilas are projected to experience mild to moderate wet conditions and central to northern (including Chapai Nawabganj, Rajshahi, Meherpur and Chuadanga zilas) upazilas to experience severe to extreme wet conditions in 2050. In monsoon season, mild wet conditions are projected to occur in most of the upazilas in Bangladesh except in some upazilas of western part (Rajshahi regions), it is likely to occur moderate to severe wet conditions in 2070. On the other hand, mild to moderate dry conditions are expected over north-eastern and southern upazilas at the same year. Under RCP4.5, severe to extreme wet conditions are projected to increase in almost all upazilas (central to southern part) for monsoon season in 2090, whereas, mild to moderate wet conditions are likely to occur over northern upazilas as the same period.
2. Monthly rainfall is expected to decrease by more than -10% in most of the months except in January, May, July, August, October and December, where July and October is likely to increase by more than 10% and in rest of the months is projected to remain the same in most of the upazilas in 2030. Seasonal rainfall is likely to decrease by more than -10% during pre-monsoon and winter season; monsoonal and post-monsoonal rainfall is expected to rise or stay normal for most of the upazilas compared to the baseline period. Annual rainfall is expected to normal in most of the upazilas under RCP4.5 in 2030.
3. During 2050, monthly rainfall is projected to remain the same in most of the months except in March, April, June, July, August and September, where March is expected to decrease by more than -10% and in rest of the months is likely to increase by more than 10% in most of the upazilas in 2050. Seasonal rainfall is likely to remain the same during pre-monsoon, post-monsoon and winter season; the monsoonal rainfall is expected to rise by more than 10% for most of the upazilas compared to the baseline period. Annual rainfall is expected to increase by more than 10% in most of the upazilas under RCP4.5 in 2050.
4. Monthly rainfall is expected to decrease by more than -10% in most of the months except in May, July, August and November, where it is likely to increase by more than 10% in most of the upazilas in 2070. Seasonal rainfall is likely to decrease by more than -10% during post-monsoon and winter season; pre-monsoonal and monsoonal rainfall is expected to rise or stay normal for most of the upazilas compared to the baseline period. Annual rainfall is expected to normal in most of the upazilas under RCP4.5 in 2070.
5. Monthly rainfall is likely to increase by more than 10% in most of the months except in March, April, June, and December, where it is likely to decrease by more than -10% in most of the upazilas in 2090. Seasonal rainfall is likely to increase by more than 10% during monsoon, post-monsoon and winter seasons; the pre-monsoonal rainfall is expected to decrease for most of the upazilas compared to the baseline period. Annual rainfall is expected to increase in most of the upazilas under RCP4.5 in 2090.
6. Monthly maximum temperature is projected to increase between 0.3 to 1.8^oC under RCP4.5. The highest increase in maximum temperature in March is projected by about 1.8^oC in the upazilas of northeast part (Sylhet region), Bangladesh. Seasonal maximum temperature is expected to rise in the range of 0.9-1.8^oC, 0.8-1.3^oC, 0.9-1.4^oC and 1.0-1.7^oC for pre-monsoon, monsoon, post-monsoon and winter season, respectively. Annual maximum temperature is projected to rise, on an average by 1.3^oC for the year 2030. Similarly, monthly minimum temperature is expected to rise by 0.2 to 1.9^oC. The highest minimum temperature in December and February is expected to increase by 1.3^oC over northwestern upazilas (Kurigram region). Seasonal minimum temperature is projected to rise in the range of 0.7 to 1.6^oC, 0.7 to 1.3^oC, 0.4 to 0.9^oC and 0.7 to 1.7^oC for pre-monsoon, monsoon, post-monsoon and winter season, respectively. Annual minimum temperature is likely to increase, on an average by 1.1^oC under RCP4.5 in 2030.   
7. The change of maximum temperature in various months at different upazilas is projected to increase by about 0.5 to 2.0^oC for the year 2050. The highest increase in maximum temperature in March is projected by about 2.0^oC over northwestern upazilas. Seasonal maximum temperature is likely to increase in the range of 1.3-1.7^oC, 1.1-1.8^oC, 1.5-2.2^oC and 1.1-1.7^oC for pre-monsoon, monsoon, post-monsoon winter season, respectively. Annual maximum temperature is projected to increase, on an average by 1.6^oC under RCP4.5. Similarly, minimum temperature change in different months are expected to rise by 1.0 to 2.6^oC over Bangladesh. The highest in minimum temperature could increase by 2.6^oC in December over western upazilas (Rajshahi region). The rate of increase in seasonal minimum temperature can occur in the range of 1.1-1.7^oC, 1.2-1.7^oC, 1.4-2.0^oC and 1.2-2.0^oC for pre-monsoon, monsoon, post-monsoon winter season, respectively and annual minimum temperature, on an average may increase by 1.6^oC for 2050.     
8. The projected increase in monthly maximum temperature could be between 1.1 to 3.2^oC for 2070. The highest increase in maximum temperature in March is expected to be about 2.9^oC over northeastern upazilas of Bangladesh. Seasonal maximum temperature is expected to rise in the range of 1.5-2.4^oC, 1.2-2.0^oC, 1.7-2.4^oC and 1.8-2.7^oC for pre-monsoon, monsoon, post-monsoon and winter season, respectively under RCP4.5 and annual maximum temperature, on an average may increase by 2.1^oC for 2070. Similarly, monthly minimum temperature is projected to increase by 1.3 to 2.9^oC. The highest in minimum temperature in December has been observed to be 2.9^oC over northeastern upazilas. Seasonal minimum temperature may vary from 1.5-2.2^oC, 1.5-2^oC, 1.8-2.7^oC and 1.8-2.5^oC for pre-monsoon, monsoon, post-monsoon and winter season, respectively and annual minimum temperature, on an average may increase by 2.1^oC for 2070.       
9. Change of maximum temperature in various months at different upazilas are projected by about 0.3-3.0^oC. The highest maximum temperature is found to be 3.0^oC in March over northwestern upazilas (Thakurgaon region) for 2090. Seasonal maximum temperature is likely to increase in the range of 1.8-3.0^oC, 1.4-2.5^oC, 1.3-2.5^oC and 0.9-1.8^oC for pre-monsoon, monsoon, post-monsoon, winter season, respectively and annual maximum temperature, on an average may be increased by 1.9^oC for 2090. On the other hand, the minimum temperature, is expected to rise by 0.9 to 2.5^oC in various months at different upazilas. The highest minimum temperature is projected to be 2.8^oC in March over northeastern upazilas (Sylhet region) for 2090. The rate of increase in seasonal minimum temperature may vary from 1.6-2.5^oC, 1.3-2.1^oC, 1.7-2.2^oC and 1.1-1.6^oC for pre-monsoon, monsoon, post-monsoon and winter season, respectively and annual minimum temperature, on an average may increase by 1.9^oC for 2090.            
10. The number of rainy days is an indicator of distribution of rainfall when viewed in conjunction with the amount of rainfall during a year. No considerable change is observed in the number of rainy days across the months except for July and October, in this two months’ rainfall is expected to be more frequent in most of the upazilas under RCP4.5. The number of rainy days is expected to increase in post-monsoon and decrease in pre-monsoon and winter season and no change in monsoon season and annual for 2030. No considerable change is expected in the projected number of rainy days for seven months, for the other five months (April, June, July, August, September) it is likely to increase for most of the upazilas. The rainy days for monsoon season is likely to increase and to remain about the same for pre-monsoon, post-monsoon and winter seasons and annual rainy days is also increased under RCP4.5 for the year 2050.
11. During 2070, the number of rainy days are expected to increase in four months (May, July, August and November) and decrease in four months (February, March, September, October) and in rest of the months to remain about the same under RCP4.5. The frequency of rainy days for pre-monsoon and monsoon season is likely to increase, and post-monsoon and winter season may decrease. Annually, the rainy days is projected to remain the same in 2070. Under RCP4.5, the rainy days are expected to be received in more days in various months except in some months (January, April, May, September and December). The rainy days of April and May are likely to decrease and in rest of the month to remain the same in 2090. The rainy days are projected to increase in monsoon, post-monsoon and winter season; whereas, in pre-monsoon season it is expected to decrease and annual frequency of rainy days is likely to increase under RCP4.5 for 2090.
12. The ratio of precipitation (P) to potential evapotranspiration (PET) provides a simple index of water supply for agriculture risk assessment. P/PET map shows in the upazilas of western part (Rajshahi regions) of Bangladesh has the highest risk for detrimental effects of lower precipitation and extended droughts as per figure for the month of June. P/PET is considerably higher than 1 may be less vulnerable to lower P and higher PET and in some cases be more vulnerable if higher P is associated with more extreme rainfall and flood. P/PPT is considerably higher than 1, have much lower risk from higher temperature and lower precipitation, because P is already in excess for the month of July-September in Bangladesh using historical data of NorESM. In monsoon season, northeastern upazilas (Moulvibazar region), where P/PET is expected by less than 1 and it may be vulnerable than the others region under RCP4.5 for the mid-century (2031-2050). Similarly, in monsoon season, northeastern upazilas (Sylhet-Moulvibazar region) where P/PET is projected by less than 1 and it may also vulnerable than the others region under RCP4.5 for the last-century (2071-2100).