Production and Service of Agrometeorological Information for the Adaptation to Climate Change in Bangladesh Dr. Sultan Ahmed (PI) Dr. A.K.M. Saiful Islam Bangladesh Agricultural Research Council (BARC) Dr. S.M. Boktiar Md. Alfi Hassan Bangladesh University of Engineering and Technology (BUET)
Background Bangladesh is an agrarian country. Agriculture is the backbone of her economy. Contribution of agriculture to national Gross Domestic Product (GDP) of the country is about 20%. Despite the importance of agriculture, the present agricultural capacity and technology of the country is hardly enough to attain self-sufficiency in food production for the fast growing population from the shrinking land resources. The information derived from hydro-meteorological and satellite observations are not readily available or analyzed to produce fruitful information for crop growth. Another challenge exists on the effective and rapid dissemination of these risks information the farmers.
Objectives Collection of local agro-meteorological data such as air temperature, precipitation, and solar radiation, etc. in all collaborative countries. Analysis of agro-meteorological variation and classification of agro-climatic zones according to crop. Changing agro-meteorological basic data into useful information such as drought index, GDD (growing degree day), crop period, etc. Maintenance and management service of agrometeorological observation system (i.e., automatic weather system) to improve reliability of agro-meteorological data.
Activities and time line Activity Collection of Secondary data (Hydro-meteorological data, crop data, soil data) Analysis of data for consistency, outliers, homogeneity and errors. Collection of existing agriculture zoning information Generation of information of agricultural applications and the frequency of agrometeorological disasters Classification of agrometeorological zones Reporting and workshop Year 1 (2012-2013) Year 2 (2013-2014) Year 3 (2014-2015) 1 2 3 4 1 2 3 4 1 2 3 4
Collection of observed meteorological data from Meteorological Department Data of 28 stations out of 34 stations of BMD used in the study which passes homogeneity and consistency test. Daily data of following variables are collected from, BMD- Rainfall Maximum and Minimum Temperature Wind speed and direction Sunshine hour Relative humidity
Observation of changes Observed data has been divided into the following two time periods each 20 years to detect the changes 1971-1990 1991-2010 Analysis has been conducted annually and seasonally Winter (Dec-Feb) Pre monsoon (Mar-May) Monsoon (Jun-Sep) Post monsoon (Oct-Nov)
Mean climatology based on records of all the 28 met stations over Bangladesh Temperature Rainfall
Mean climatology based on records of all the 28 met stations over Bangladesh Temperature Maximum Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 1971-1990 25.5 27.9 31.5 33.0 32.8 31.7 30.8 31.2 31.4 31.3 29.4 26.3 1991-2010 25.0 28.2 31.9 33.5 33.3 32.3 31.6 31.9 32.0 31.8 29.7 26.6 Temperature Minimum Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 1971-1990 12.7 15.1 19.5 23.2 24.3 25.5 25.5 25.6 25.3 23.6 19.2 14.2 1991-2010 12.4 15.5 20.0 23.4 24.7 25.7 25.9 25.9 25.5 23.8 19.2 14.4 Rainfall Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 1971-1990 10 23 51 118 287 462 527 413 325 183 40 10 1991-2010 12 25 54 101 289 464 516 409 336 199 34 9
Mean climatology based on records of all the 28 met stations over Bangladesh Humidity Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 1971-1990 70.8 68.5 68.6 73.5 78.7 84.5 86.9 85.7 85.1 80.9 75.2 72.6 1991-2010 72.4 69.9 69.8 74.4 78.7 84.2 86.1 85.3 84.5 81.2 75.4 72.9 Wind Speed Wind Speed Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 1971-1990 2.7 3.1 3.8 5.1 4.6 4.7 4.5 4.4 3.5 2.7 2.5 2.5 1991-2010 2.6 2.9 3.5 3.9 4.1 4.0 3.9 3.6 3.1 2.5 2.1 2.2
Changes of Meteorological Drought of Bangladesh
Detection of Drought using SPI (Standard Precipitation Index) Drought events have severe impact on country s agricultural economy in past years. Between 1960 and 1991, droughts occurred in Bangladesh 19 times. Very severe droughts hit the country in 1951, 1961, 1975, 1979, 1981, 1982, 1984, and 1989. The Standard Precipitation Index (SPI) is a very useful to predict meteorological drought over a region
Meteorological Drought Assessment Standardized Precipitation Index (SPI), a tool derived by McKee et al. (1993), a measure of meteorological drought has been calculated from the available rainfall data. Mathematically, SPI is calculated based on the following equation SPI ( Xi Xm) where, Xi is monthly rainfall record of the station; Xm is rainfall mean; and σ is the standard deviation
SPI based Drought Severity Index Range Condition SPI -2 Extremely dry -2 < SPI -1.5 Severely dry -1.5 < SPI -1 Moderately dry -1 < SPI 1 Near normal 1 < SPI 1.5 Moderately wet 1.5 < SPI 2 Severely wet SPI 2 Extremely wet
Calculation of SPI 5 time period has considered to calculate SPI. 1- month (monthly SPI) 3- month (Seasonal SPI) 6-month (Short time SPI) 9 month (Medium time SPI) 12 month (Long term SPI) SPI is calculated both temporally and spatially.
Number of extreme drought events during 1971-1990 and 1991-2010 3-month SPI Extreme Droughts during Rabi season
Number of severe drought events during 1971-1990 and 1991-2010 3-month SPI Severe Droughts during Rabi season
Number of moderate drought events during 1971-1990 and 1991-2010 3-month SPI Moderate Droughts during Rabi season
Changes of drought severity Seasonal SPI or 3-month SPI is useful to understand the soil moisture condition of an area. Considering 3-month SPI it has found that, frequency of extreme drought increased in the north western part of Bangladesh. Using SPI-6 month, long term seasonal extreme drought increased rapidly from 1980s to 2000s time period, especially over entire north-western region of Bangladesh.
Changes of Potential Evapo-transpiration (PET)
Potential evaporation or potential evapotranspiration (PET) PET is defined as the amount of evaporation that would occur if a sufficient water source were available. FAO Penman-Monteith method to estimate ET 0 can be derived as follow: ETo = reference evapotranspiration [mm day-1] Rn = net radiation at the crop surface [MJ m-2 day-1] G = soil heat flux density [MJ m-2 day-1] T = mean daily air temperature at 2 m height [ C] u2 = wind speed at 2 m height [m s-1] es = saturation vapour pressure [kpa] ea = actual vapour pressure [kpa] es-ea = saturation vapour pressure deficit [kpa] Δ = slope of vapour pressure curve [kpa C-1] γ = psychrometric constant [kpa C-1]
Evapotranspiration ETo expresses the evaporating power of the atmosphere at a specific location and time of the year and does not consider the crop characteristics and soil factors. A software called CROPWAT 8.0 which is developed by Water Resources Development and Management Service of FAO. Six climatic variables namely, Maximum temperature, minimum temperature, sunshine hour, humidity, wind speed are used to calculate potential evapotranspiration.
Monthly PET over Bangladesh during 1971-1990 and 1991-2000 5.0 4.5 4.0 1971 to 1990 1991 to 2010 3.5 3.0 2.5 2.0 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Monthly PET during Kharif-I season PET (1971-1990) PET (1991-2010)
Monthly PET during Kharif-II season PET (1971-1990) PET (1991-2010)
Monthly PET during Rabi season PET (1971-1990) PET (1991-2010)
Spatial patterns of the changes of PET Kharif-I, Kharif-II and Rabi (left to right)
Changes of PET It can be suggested that ET has dropped at Kharif I season in recent climate. South eastern region of Bangladesh shows a notable decrease in ET especially at Meherpur District Kharif II season experiencing lesser amount of ET during 1991-2000. ET reduces more at Western part than eastern part of the country. On the other hand during Rabi season, ET does not seem significant change around the country.
Classification of changes based on Agro- Ecological Zones (AEZ) of Bangladesh
% Changes of PET for different AEZ Agro-ecological Zone Rabi (%) Kharif-I (%) Kharif-II (%) Old Himalayan Piedmont Plain 5.6 3.9 5.8 Active Tista Floodplain -4.4-5.3 1.9 Tista Meander Floodplain -3.4-3.8 2.0 Karatoya-Bangali Floodplain -7.9-5.6-1.1 Lower Atrai Basin -6.5-6.0-2.3 Lower Purnabhaba Floodplain 2.8 2.5 1.0 Active Brahmaputra-Jamuna Floodplain -6.0-4.4 0.8 Young Brahmaputra and Jamuna Floodplains 0.6 0.1 2.4 Old Brahmaputra Floodplain 0.3-1.7 2.2 Active Ganges Floodplain -3.3-2.8 0.9 High Ganges River Floodplain -14.2-11.6-3.2 Low Ganges River Floodplain -4.8-4.2 3.5 Ganges Tidal Floodplain 1.8 6.8 3.1 Gopalganj-Khulna Bils 0.5 1.8 8.5 Arial Bil 6.7 5.7 8.6 Middle Meghna River Floodplain -2.3-3.0-1.0 Lower Meghna River Floodplain 2.3 0.4 0.7 Young Meghna Estuarine Floodplain -3.2-1.9 2.4 Old Meghna Estuarine Floodplain -0.1-2.4 1.5 Eastern Surma-Kusiyara Floodplain 5.2 3.5 6.3 Sylhet Basin 0.5-6.0 1.4 Northern and Eastern Piedmont Plains -2.9-4.1 2.1 Chittagong Coastal Plain -6.6-3.3-2.7 St. Martin's Coral Island -13.2-9.0-11.1 Level Barind Tract -3.6-3.3 0.4 High Barind Tract 0.7 0.7-0.7 North-Eastern Barind Tract -4.2-4.1 1.7 Madhupur Tract 6.2 4.7 3.1 Northern and Eastern Hills 6.8 6.5 8.1 Akhaura Terrace -6.1-7.9-0.5 Reserved Forest 20.9 17.9 19.2
Changes of Growing Degree Days (GDD)
Growing Degree Days Growing degree days (GDD), also called growing degree units (GDUs), are a measure of heat accumulation used by horticulturists, gardeners, and farmers to predict plant and pest development rates such as the date that a flower will bloom or a crop reach maturity. Where, Tmax and Tmin is the daily maximum and minimum temperatures compared to a base temperature, T base, (usually 10 C)
Crop wise Growing Degree Days, GDD Crop Name Growing Seasons Sowing /Planting Base Period ( C) Rice (T - Aus) Kharif - I Mid-March to Mid-April 10 Rice (T Aman) Kharif I Late-June to Early September Rice (Boro) Rabi December to Mid- 10 February Maize Rabi October to November 10 Potato Rabi Mid-September to Mid- November Mustard Rabi November to early 10 December Wheat Rabi November 5.5 10 8
GDD for Kharif I season
GDD for Kharif II season
GDD for Rabi Season
Changes of GDD GDD for Aman rice has been increased. GDD of Aus rice has been not significantly changes though its spatial variability has been increased. GDD of Boro has been increased. Changes of climate influences GDD for different seasons of the year.
Changes of GDD of rice for different AEZ Agro-ecological Zone Boro Kharif II Kharif I Change Change Change Old Himalayan Piedmont Plain 0.5 61.4 53.9 Active Tista Floodplain -6.4 64.1 12.9 Tista Meander Floodplain -3.9 60.5 25.7 Karatoya-Bangali Floodplain 8.3 64.1 53.5 Lower Atrai Basin 3.0 66.7 68.4 Lower Purnabhaba Floodplain -1.5 64.5 70.6 Active Brahmaputra-Jamuna Floodplain 4.0 72.5 54.8 Young Brahmaputra and Jamuna Floodplains 28.0 90.8 115.7 Old Brahmaputra Floodplain 52.1 134.1 179.9 Active Ganges Floodplain 18.5 64.5 81.7 High Ganges River Floodplain 12.5 59.4 63.7 Low Ganges River Floodplain 25.4 61.8 72.0 Ganges Tidal Floodplain 40.6 49.8 66.9 Gopalganj-Khulna Bils 35.1 60.9 65.8 Arial Bil 19.9 54.4 75.8 Middle Meghna River Floodplain 28.6 67.5 90.4 Lower Meghna River Floodplain 41.2 86.7 119.3 Young Meghna Estuarine Floodplain 9.4 48.4 72.5 Old Meghna Estuarine Floodplain 25.8 63.3 87.3 Eastern Surma-Kusiyara Floodplain 65.4 100.6 100.3 Sylhet Basin 56.1 98.6 123.0 Northern and Eastern Piedmont Plains 52.8 103.9 127.5 Chittagong Coastal Plain -16.8-14.1-25.5 St. Martin's Coral Island 69.6-2.6 57.9 Level Barind Tract -1.4 62.0 51.4 High Barind Tract -2.9 65.1 79.7 North-Eastern Barind Tract -5.2 56.7 18.3 Madhupur Tract 29.2 91.0 123.4 Northern and Eastern Hills -1.9 26.1 25.1 Akhaura Terrace 27.8 56.1 77.4 Reserved Forest -5.7 19.9 21.7
Changes of Extreme climate
Changes of Extreme climate The joint Expert Team (ET) on Climate Change Detection and Indices (ETCCDI) has recognized a suite of 27 core climate change indices which were derived from daily precipitation and temperature data. Extreme indices of ERCCDI that are used in this study. Out of that list, a number of standard indicators are chosen to quantify changes of extreme climate which are listed below- Consecutive wet days, CWD Consecutive dry days, CDD Number of days when Rain > 100mm, RX100 Changes of Minimum temperature Changes of maximum temperature Changes of Diurnal ranges of temperature
Consecutive Wet Days
Consecutive Dry Days
Number of rainy days when rainfall is more than 100mm
Minimum temperature during December
Maximum temperature during December
Maximum temperature during April (summer)
Changes of Extremes It has been found that both CDD and CWD days has been increased. Hence, monsoon becomes more wetter and dry season becomes more drier. Chances of flooding has increased. Both maximum and minimum temperature has been increase during December which pose adverse impact on clod loving crops. Summer becomes more hotter than past.
Changes of minimum temperature during December for different AEZ Agro-ecological Zone 1971-1990 1991-2000 Change Old Himalayan Piedmont Plain 9.1 9.4 0.3 Active Tista Floodplain 9.6 9.8 0.2 Tista Meander Floodplain 9.5 9.7 0.3 Karatoya-Bangali Floodplain 9.9 10.1 0.2 Lower Atrai Basin 9.3 9.5 0.3 Lower Purnabhaba Floodplain 9.1 9.4 0.3 Active Brahmaputra-Jamuna Floodplain 10.0 10.1 0.1 Young Brahmaputra and Jamuna Floodplains 10.4 10.5 0.1 Old Brahmaputra Floodplain 10.2 10.4 0.2 Active Ganges Floodplain 9.9 10.3 0.4 High Ganges River Floodplain 9.2 9.3 0.1 Low Ganges River Floodplain 10.3 10.7 0.3 Ganges Tidal Floodplain 10.8 11.2 0.4 Gopalganj-Khulna Bils 10.6 11.0 0.4 Arial Bil 10.9 11.4 0.5 Middle Meghna River Floodplain 10.8 11.3 0.5 Lower Meghna River Floodplain 11.1 12.0 0.8 Young Meghna Estuarine Floodplain 11.6 11.6 0.0 Old Meghna Estuarine Floodplain 10.7 11.0 0.3 Eastern Surma-Kusiyara Floodplain 9.8 10.8 1.0 Sylhet Basin 9.8 10.3 0.5 Northern and Eastern Piedmont Plains 9.8 10.2 0.4 Chittagong Coastal Plain 12.6 12.4-0.2 Level Barind Tract 9.4 9.7 0.3 High Barind Tract 9.1 9.2 0.2 North-Eastern Barind Tract 9.5 9.8 0.3 Madhupur Tract 10.5 10.7 0.2 Northern and Eastern Hills 12.2 11.8-0.4 Akhaura Terrace 10.3 10.5 0.2
Changes of maximum temperature for different AEZ during April Agro-ecological Zone 1971-1990 1991-2000 Change Old Himalayan Piedmont Plain 37.92 37.0-0.9 Active Tista Floodplain 37.43 36.3-1.1 Tista Meander Floodplain 37.93 36.8-1.2 Karatoya-Bangali Floodplain 39.06 37.8-1.2 Lower Atrai Basin 39.78 38.6-1.1 Lower Purnabhaba Floodplain 39.12 38.0-1.1 Active Brahmaputra-Jamuna Floodplain 38.30 37.2-1.1 Young Brahmaputra and Jamuna Floodplains 37.55 37.0-0.5 Old Brahmaputra Floodplain 37.01 36.4-0.6 Active Ganges Floodplain 38.39 38.1-0.3 High Ganges River Floodplain 39.72 39.1-0.6 Low Ganges River Floodplain 37.77 37.7 0.0 Ganges Tidal Floodplain 36.57 37.0 0.4 Gopalganj-Khulna Bils 37.18 37.3 0.1 Arial Bil 36.92 36.9 0.0 Middle Meghna River Floodplain 36.02 36.2 0.1 Lower Meghna River Floodplain 35.67 35.8 0.1 Young Meghna Estuarine Floodplain 35.25 35.8 0.5 Old Meghna Estuarine Floodplain 35.72 36.1 0.4 Eastern Surma-Kusiyara Floodplain 35.08 35.7 0.6 Sylhet Basin 35.94 36.0 0.1 Northern and Eastern Piedmont Plains 36.05 36.1 0.0 Chittagong Coastal Plain 34.52 35.5 1.0 Level Barind Tract 39.08 37.7-1.3 High Barind Tract 39.82 38.8-1.0 North-Eastern Barind Tract 38.06 36.8-1.3 Madhupur Tract 37.38 36.9-0.5 Northern and Eastern Hills 34.91 35.9 1.0 Akhaura Terrace 37.92 37.0-0.9
Climate Zoning for Bangladesh
Climate Zoning Two climatic periods are considered in the study; 1971-1990 as prior climate and 1991-2010 as current climate. Three meteorological variables, rainfall, temperature maximum (TMAX) and temperature minimum (TMIN) are selected for the study. To identify the variability of different climatic parameter, initial K-mean clustering has been made for each variable.
K-mean clustering k-means clustering is a method of vector quantization, originally from signal processing, that is popular for cluster analysis in data mining. k- means clustering aims to partition n observations into k clusters in which each observation belongs to the cluster with the nearest mean, serving as a prototype of the cluster. This results in a partitioning of the data space into Voronoi cells.
Example of k-mean clustering
K-mean clustering For the K-mean clustering it is necessary to determine the optimum cluster size for future analysis. To identify optimized number of cluster, caliski criterion and weighted sum error (SSE) has been used for all the individual approach of the K-mean clustering in this summer. After analyzing, TMAX, TMIN and rainfall of current climate, it has found that optimum number of K- mean is 3 over Bangladesh.
K-mean Clustering K-mean clustering for rainfall has been applied for current and past climate. The clustering has been conducted in monthly scale for 20 years time period. Firstly optimized climate zones for rainfall, temperature maximum and temperature minimum have been identified and characterized. Zonal statistic of each zones are identified to define zone boundaries. For single variable zoning, static K-mean algorithm applied to have detail and specific picture of change of climate.
Properties for each climate zones based on Precipitation, Max Temp. and Min Temp. Zone Name Variable Unit Winter Summer Monsoon Post- Monsoon Annual ZP1 precipitation mm 39 ± 9 364 ± 124 1397 ± 263 206 ± 71 2000 ± 422 ZP2 precipitation mm 47 ± 6 671 ± 152 1806 ± 329 218 ± 12 2750 ± 503 ZP3 precipitation mm 51 ± 12 512 ± 107 2250 ± 473 277 ± 57 3100 ± 650 ZTX0 Max. Temp. deg <25.5 <33.0 <31.8 <30.7 <30.5 ZTX1 Max. Temp. deg 26.3 ± 0.8 33.9 ± 0.9 32.4 ± 0.6 30.8 ± 0.1 30.8 ± 0.3 ZTX2 Max. Temp. deg 25.9 ± 0.7 31.9 ± 0.7 32 ± 0.5 30.7 ± 0 30.2 ± 0.3 ZTX3 Max. Temp. deg 27.1 ± 0.6 32.7 ± 0.9 31.6 ± 0.5 30.8 ± 0.1 30.5 ± 0.3 ZTX4 Max. Temp. deg >27.7 >33.6 >32.1 >30.9 >30.8 ZTN0 Min. Temp. deg <12 <20.5 <25.5 <20 <19.5 ZTN1 Min. Temp. 12.9 ± 1 21.8 ± 1.2 25.7 ± 0.2 20.9 ± 0.9 20.3 ± 0.8 ZTN2 Min. Temp. deg 12.7 ± 1 20.8 ± 1 25.2 ± 0.4 20.3 ± 0.8 19.7 ± 0.7 ZTN3 Min. Temp. deg 14.8 ± 1.2 23 ± 0.7 25.5 ± 0.5 21.9 ± 0.5 21.3 ± 0.5 ZTN4 Min. Temp. deg >16 >23.5 >26 >22.5 >21.5
Optimized number of K-mean cluster over Bangladesh for different variables. (a) TMAX; TMIN (b) Rainfall
Change of climate zones over Bangladesh for different climate variables
Seasonal Climate zones of Bangladesh and their observed shift.
Combined Climate zones of Bangladesh and their observed shift.
District wise climate zone
Result Dissemination and Publications
An national workshop was held at BARC on March 10 2014 to share the project activities with NARS scientist and other related stakeholders.
Dr. S M Bokhtiar, is presenting 2113-2014 progress report at PI meeting held on June 03-07, 2014 at Ulaabaater, Mongolia
Fig.1. Presentation of project activities in a national workshop which was held at BARC on 10 March 2014
Presentation was made in the Annual Evaluation meeting of AFACI Projects in Bangladesh on 6 June 2015 at BARC, Bangladesh
Publications Book Boktia, S.M., Ahmed, S., Islam, A.K.M.S., Hasan, M.A. (2015) Retrosp ective Analysis of Agro-meteorological Information in Bangladesh, Bangladesh Agricultural Research Council (BARC), Dhaka, ISBN: 97 8-984-500-026-0. Articles in Conference proceedings Hasan, M. A., Islam, A.K.M.S. and Boktiar, S.M. (2014) Changes of R eference Evapotranspiration (ET0) in recent decades over Bangla desh, Proceedings of the 2nd International Conference on Advan ces in Civil Engineering 2014 (ICACE-2014), 26 28 December, 2014, CUET, Chittagong, Bangladesh, Vol. 1, pp. 772-777.
Conclusions This study has successful collect local hydro-meteorological data such as air temperature, precipitation, and solar radiation, etc. of Bangladesh from all available observatories for the last 40 years. This study has analyzed agro-meteorological basic data into useful information such as drought index, evaporation, growing degree day etc. Changes of these indicators were quantified and results have been presented in the present agro-ecological zones of the country. Using k-mean clustering algorithm, climate zoning has been performed for the country considering three major weather variables (precipitation, maximum temperature and minimum temperature). Finally, results were disseminated through national level workshop, annual meeting, country reports and articles published in conferences.
Recommendations/Future Plan Comprehensive drought risk assessment: In this study, only meteorological drought has been assessed using observed rainfall information. However, agricultural droughts based on soil moisture information of the field can be combined with meteorological drought to get more comprehensive picture of drought. Agro-climatic zoning: It was also found important to develop an agro-climatic zone for Bangladesh. At present results has been presented in the agro-ecological zones of Bangladesh. Using k-mean algorithm, climate-zoning maps for the country have been developed. However, based on location of similar climatic zones, agro-ecological zones can be further merged and developed into a common agro-climatic zone. Web dissemination portal: Results of this study should be disseminated to the policy makers, researchers and major stakeholders of the nations. In this context, web based information portal can be developed to dissemination major findings of this study.