INTERNATIONAL JOURNAL OF GEOMATICS AND GEOSCIENCES Volume 3, No 3, 2013

Transcription

1 INTERNATIONAL JOURNAL OF GEOMATICS AND GEOSCIENCES Volume 3, No 3, 2013 Copyright by the authors - Licensee IPA- Under Creative Commons license 3.0 Research article ISSN Monitoring meteorological and agricultural drought dynamics in Barind region Bangladesh using standard precipitation index and Markov chain model 1, Sayedur Rahman M 2, Saadat A.H.M 1 1- Department of Environmental Sciences, Jahangirnagar University, Bangladesh 2- Department of Statistics, University of Rajshahi, Bangladesh kallal1022@gmail.com ABSTRACT The Barind region of Bangladesh suffers from frequent drought due to poor and delayed rainfall in pre monsoon and monsoon season. In the present study details analysis of rainfall data has been carried out for the years Two well known drought indices, Standard precipitation index (SPI) has been used to evaluate the precipitation deficit, and Markov chain model used to quantify the drought in agricultural extent. Details analysis of spatiotemporal drought dynamics during Pre-Kharif (pre-monsoon) and Kharif (monsoon) have been carried out through drought indices map generated in a GIS (Geographical Information System) environment. Analysis and interpretation of these maps evaluate that meteorological drought is not always correspond to agricultural drought and a chronic agricultural drought can be occur because of a limited rainfall deficit. The study also identified that there is a statistically significant temporal correlation between SPI and Markov chain drought indices but spatial correlation is poor in pre monsoon season. In the Barind region meteorological drought shows a similar pattern in pre monsoon season but in monsoon season rainfall deficits shifts its position time to time and in certain pockets it is more frequent. In agricultural sphere, the Markov chain shows a very chronic pattern during pre-monsoon season in the entire area but in monsoon it behaves like meteorological drought.. Keywords: SPI, Markov chain model, GIS, Barind region, drought. 1. Introduction Drought is a reoccurring phenomenon in the northwestern part of Bangladesh. Though the drought has attracted less scientific attraction than flood or cyclone, several authors found that the impact of drought can be more defenseless than flood and cyclone (e.g. Shahid and Behrawan, 2008; Shahid, 2008). This is also evident by the statistics that, in the loss of 1978/79 drought was greater than the loss of flood in 1974 (Paul, 1998). Moreover, Rice production losses in the drought of 1982 were 50 percent more than the losses occur due to the flood in the same year (Ramsey et al., 2007). Furthermore, in the 1997 drought the country reduced 1 million tons of food grains of which 0.6 million tons were T. aman (Ramsey et al., 2007). On the global level, impact of natural hazards and disasters are staggering. In Bangladesh, the major natural hazards are also in line with global patterns. In the context of global warming, most of the climatic models project a decrease in precipitation in dry season and an increase during monsoon in south Asia (Christensen et al., 2007; Shahid and Behrawan, 2008). This will cause a cruel combination of more extreme floods and droughts in the region. Due to the Submitted on January 2013 published on March

2 land use changes within the country and in neighboring country, Bangladesh has already showed an increased frequency of droughts in recent years (Shahid and Behrawan, 2008). Concern among scientists has grown on changes of precipitation and frequent occurrence of droughts in Bangladesh. Therefore, it is necessary to identify different kinds of drought (e.g. meteorological, Agricultural etc.) for taking proper management plan. Moreover, Drought is a highly misused word because there is no universally accepted definition of drought. Several authors defined drought on the basis of meteorological, agricultural, hydrological and socio economic impact (Darcup et al., 1980; Wilhite and Glantz, 1985; Ramsey et al., 2007) and also tries to linking up the various types of drought. Meteorological drought is directly related to the weather parameter rainfall, but agricultural drought in is the consequence of meteorological drought (Ramsey et al., 2007). Agricultural drought is mainly caused by the soil water deficiency which is occurred because of meteorological derivation. Standard precipitation index (SPI) is the most widely used meteorological drought index. The advantages of SPI are that it can be calculated for a variety of time scales. This versatility allows SPI to monitor short-term water supplies, such as soil moisture which is important for agricultural production, and long-term water resources, such as groundwater supplies, stream flow, and lake and reservoir levels. Soil moisture conditions respond to precipitation anomalies on a relatively short scale (Mishra and Singh 2010). SPI has been used for studying different aspects of droughts, for example, forecasting (Mishra et al., 2007), frequency analysis (Mishra et al., 2009), spatio temporal analysis (Mishra and Singh, 2009; Shahid, 2008; Loukas and Vasiliades, 2004; Bastini, 2011) and climate impact studies (Mishra and Singh, 2009; Loukas et al., 2008), monitoring ground water drought (Shahid and Hazarika, 2010) assessing drought risk (Shahid and Behrawn, 2008). On the other hand Markov chain model is a widely used agricultural drought index in the Indian sub continent. It has been used for determining drought proneness (Banik et al., 2002), Analyzing rainfall drought correlation (Alam et al., 2011) Analyzing agricultural drought (Bimah et al., 2005) and predicting critical wet and dry spell (Ochola and Kerkides, 2003). Several studies also cited that the reliability of meteorological persistence can be best described through Markov chain model of proper order (Rahman, 1999a&b), Moreover, Markov chain model is found to be promising in simulating the length of the longest dry and wet spells and largest rainfall amount during monsoon (Sharma, 1996; Biamah, et. al., 2005). The drought measuring parameters are not linearly related to one another, these drought indices often have little correlation among themselves (Bhuyian et al., 2006). Therefore it is quite common that when one drought index identifies drought at a particular place, another drought index indicates normal condition at the same time and place (Bhuyian et al., 2006). Monitoring drought is challenging because of the large spatial and temporal variability of drought events. Monitoring techniques need to be adapted to capture the time location and sector specific characteristics of drought. Moreover, use of a particular specific index has often been demonstrated to be inadequate for completely representing this complex phenomenon (Heim, 2002; Steinemann et al., 2005; Quiring, 2009). In the present study, detailed analysis of seasonal drought dynamics has been carried out to identify spatiotemporal drought patterns in meteorological and agricultural spheres. Time-series analysis of drought and spatial maps of the Barind region have been generated in a Geographic Information Systems (GIS) using SPI and Markov chain model drought indices. Since aquifer-recharge, agricultural activities, and ecological changes are controlled by rainfall, the present analysis has been focused on drought during the monsoon and the nonmonsoon periods. Standardized Precipitation Index (SPI) has been used to monitor meteorological drought and Markov chain model to monitor and analyzing agricultural 512

3 drought. Finally, relative drought dynamics in meteorological and agricultural sphere of the Barind region have been compared. 2. Climatic condition of Barind region Barind Tract is the largest Pleistocene physiographic unit of the Bengal Basin which covers an area of about 7,770 sq km. Three distinct seasons can be recognized in area from climatic point of view: (1) the dry winter season from December to February, (2) the pre-monsoon hot summer season from March to May, and (3) the rainy monsoon season which lasts from June to October (Shahid, 2011). Climatically, the region is characterized by high fluctuating rainfall and the ratio of dry to rainy months is found to be highest in Bangladesh. It is observed that annual average rainfall vary between 1,400 and 1900 mm and the seasonal distribution of rainfall shows that more than 90 % rainfall occurs during May to October (Shahid, 2011). Although the whole region has long cool winter, the maximum number of days with temperature above 40 C is observed in summer (Rahman, 1999a). Geologically Barind tract belongs to an old alluvial formation which is usually composed of massive argillaceous beds of pale reddish brown color that often turns yellowish on weathering. Lime nodules and pisolitic ferruginous concretions occur throughout the soil. Locally the soils are rich in lime. Soils are deficient in nitrogen and phosphorus (Moslehuddin et al., 2008). The daily rainfall data of 12 rainfall stations were available from the year 1971 to 2008 in Bangladesh Water Development Board (BWDB). Location of the rainfall stations are given in figure 1. Figure 1: Location of the rainfall stations in the study area 513

4 3. Drought Indices 3.1. Monitoring of meteorological drought McKee et al. (1993, 1995) proposed Standardized Precipitation Index to assess anomalous and extreme precipitation. Since precipitation data are mostly skewed, in order to compute SPI, precipitation data are normalized using gamma function. SPI is based on the probability of precipitation for any desired time scale. The SPI is computed by dividing the difference between the normalised seasonal precipitation and its long-term seasonal mean by the standard deviation. Thus Where, Xij is the seasonal precipitation at the ith rain gauge station and jth observation, Xim the long-term seasonal mean and s is its standard deviation. Five classes of SPI as shown in Table 1 are used in the present study. 3.2 Monitoring of agricultural drought Several authors have found that the sequences in daily rainfall occurrences can be described by a simple Markov chain model. Additional evidence to indicate the feasibility of using a Markov chain model has been presented by Rahman (1999 a&b), Banik et al. (2002), Borkatulla (2007) and Alam et al., (2011). The theory of Markov chain is described below: Let X 0,X 1,X 2,,X n, be random variables distributed identically and taking only two values, namely 0 and 1, with probability one, i.e., 0 if the nth week is dry X n = 1 if the nth week is wet Firstly, it may be assume that, P ( X n+1 = x n+1 X n = x n,x n-1 = x n-1,,x 0 = x 0 ) = P( X n+1 = x n+1 X n = x n ) where x 0,x 1,.,x n+1 { 0,1 }. In other words, it is assumed that probability of wetness of any week depends only on whether the previous week was wet or dry. Given the event on previous week, the probability of wetness is assumed independent of further preceding weeks. So, the stochastic process {X n, n = 0, 1, 2.} is a Markov chain. Considering the transition matrix as; P P P01 P 11 where P ij = P(X 1 = j X 0 = i) i,j = 0,1. Note P 00 +P 01 = 1 and P 10 +P 11 = 1 P 11 gives the probability of a week to be wet given that previous week was wet also. When P 11 is large, the chance of wet weeks is also large. But only a small value of P 11 may not indicate high drought proneness. In this case, large value of P 01 implies a large number of short wet spells which can prevent occurrence of drought. 514

5 Hence, an index of drought proneness may be defined as DI = P 11 P 01 (2) This index of drought proneness is bounded by zero and one. Higher the value of DI, lower will be the degree of drought proneness. Table 1: Index of Drought Proneness Drought Classes SPI Markov Chain Model Extreme Drought <-2.00 >0.125 Severe Drought <-1.5 >0.185 Moderate Drought <-1.00 >0.235 Mild Drought <-.00 >0.310 No drought >0.00 > Mapping of meteorological and agricultural drought For mapping of spatial extent of meteorological and agricultural droughts from point data, a kriging interpolation method is used. In this study, Inverse distance weighting (IDW) of Geostatistics tools has been used for mapping the spatial extent of meteorological and agricultural droughts because it is a quick deterministic interpolator that is accurate. Moreover there is a very diminutive decision to make regarding model parameter. Furthermore, Geostatistics is based on the theory of regionalized variables are increasingly preferred because it allows the capitalization of spatial correlation between neighboring observations to predict attribute values at not sampled locations (Goovaerts, 2000). 4. Result and Discussion The use of two or more drought indices provide and understanding of drought characteristics and probability of re-occurrences of drought easily (Mpelasoka et al., 2008). It discussed the critical stages and can also able to identify real drought events. In the present study both SPI and Markov chain model have been used to observe drought event during the period For meteorological drought rainfall from March to May considered as Pre-monsoon season or and June to October has considered as monsoon (Shahid, 2011). On the other hand Markov chain model is a threshold based model. In the present study weekly 20 mm rainfall considered as threshold value (Alam et al., 2012) and occurrence of agricultural drought has been calculated by considering 13 and 21 standard weeks for Pre-monsoon and monsoon seasons respectively. 4.1 Monitoring Meteorological drought Meteorological drought is a very common phenomenon in the north-western part of Bangladesh (Shahid, 2008; Shaid and Hazarika, 2010). This statement is also supported by the present study. Chronic to mild all classes of meteorological drought occurred during the period in Pre-monsoon season in discrete pocket of the study area. The time series of average SPI value in both Pre-monsoon and monsoon seasons are given in figure 2. The figure 2 evident that in both Pre-Kharif and Kharif seasons the area is affected by different magnitude of drought in about 19 times. But the different parts of the area are 515

6 affected in different years. The figure also suggests that the frequency of drought has increased in 1990s and recent decays compare to 1970s and 1980s. And it is clear that the study area is affected by drought almost ones in two years. The major drought events in Premonsoon season (during ) are given figure 3. It is evident that a north-eastern pocket of the study area is more vulnerable to meteorological drought during Pre-monsoon season. Moreover, in all the years the category of drought occurrence showed similar patterns within a few exceptions. The incidences of meteorological drought also partially support the findings of Shahid (2008). Where he found that the major four meteorological droughts in pre-monsoon was in 1982, 1989, 1992 and 1995 in the western part of Bangladesh. In case of monsoon period, meteorological drought occurred in different pockets in different years. Unlike pre-monsoon it does not show any similar pattern. Major ten monsoon drought events have been given in the figure 4. Chronic meteorological drought affects the area in the year 1982, 1994 and Among these drought events the drought of 2010 was most devastating. In addition to this severe drought occur almost all the year cited in the figure 4. However, moderate to mild drought occurred almost every year in some discrete pocket of the study area during Figure 2: Time series of average meteorological drought index (a) Pre-monsoon season (b) monsoon season The drought event of monsoon season also partially supported the result of Shahid (2008) in which he identified 1982, 1989, 1992 and 1995 as the most devastating drought events in the western part of Bangladesh. During the just within a span five years monsoon drought affect the Barind region four times. Shahid (2010) found that the monsoon rainfall in the northwestern part of Bangladesh has increased significantly during the time period So, this drought event is contradicted with the findings of Shahid (2010). Other results disagree with the findings of Shahid (2008, 2010) as he stated that there is no significant correlation between rainfall and meteorological drought but all the definition of 516

7 meteorological drought stated that lack of precipitation is the principle cause of meteorological drought (Ramsey et al., 2007) Figure 3: Major meteorological drought events in Pre-monsoon season Figure 4: Major meteorological drought events in monsoon season 4.2 Monitoring Agricultural drought The time series of agricultural drought occurrence is given in the figure 5. Figure 5(a) explained that during Pre- monsoon growing season almost all the year affected by chronic to moderate agricultural drought. And the occurrence of agricultural drought during Kharif growing season is quite lower. During around 5 to 6 time s monsoon agricultural drought affect the area. But drought occurs almost every year in some discrete pockets of 517

8 Barind region. This is may be because of most of the rainfall in this region occurs during the month June to October and during November to May the region enjoys a very little rainfall (Rahman, 2000). (a) (b) Figure 5: Time series of average agricultural drought index (a) Pre-monsoon season (b) Monsoon season Figure 6: Major agricultural drought events in Pre-moonsoon season Figure 7: Major agricultural drought events in monsoon season 518

9 During pre- monsoon season chronic drought occur almost every year in any parts of the study area. From the figure 6 it is evident that during there were at least 5 years (1972, 1979, 1995, 2005 and 2008) when the entire study area has been affected by the chronic drought. This significance of chronic drought resulted from very low rainfall (almost zero) and the high temperature which influence the moisture holding capacity of the soil. This chronic agricultural drought during Pre-Kharif growing season increase the groundwater irrigation demand in Boro rice field which shares almost 70% of the total rice production of Bangladesh. But in case of Barind region lack of rain or chronic drought compel the farmers to not to seed their Boro rice and around 85% of the rain remain fallow during this season in Barind region (Alam et al., 2011; Sadaat el al., 2009). During Kharif growing season six major droughts year were identified throughout time period And four droughts years were identified during the last decay of twenty century (figure 7). The agricultural drought of 1994 also supported the statement of Shahid (2008) and Shahid and Behrawan (2008) where they stated in the drought of led to a decrease in rice and wheat production of ton. Moreover, Dey et al., (2011) marked 1994 as the driest year in the contemporary periods which lead to devastating impact on the northwestern part of Bangladesh. The drought events of 1997 support the declaration of Ramsey et al., (2007) affirming that the drought of 1997 has lead to short fall of1 million tons of food grains of which 0.6 million tons were T. aman. Along with this agricultural droughts occur in some discrete pockets of the Barind region almost every year. Ramsey et al., (2007) confirmed that around 2.32 million hectare of T.aman rice crops damages due to agricultural drought every year. On the other hand Alam et al., (2011) and Rahman (2000) indicate this amount is to millions of hectares. 4.3 Relative drought dynamics The coefficients of correlation between meteorological drought index and agricultural drought index in twelve rainfalls station in the studied time frame ( ) for both Premonsoon and monsoon seasons are also studied. The result from Pearson correlation indicates that all of the rainfall stations, the meteorological and agricultural drought indices are statistically significant except in two stations namely Nithpur and Badalgachi. However the degree of correlation varied among different rainfall stations (r=0.472 to 0.813). SPI and Markov chain drought index demonstrated weakest correlation especially in wet season (r=0.139 to 0.801). This may be attributed to the fact the degree of agricultural drought not only related to occurrence of rainfall but other conditions such as soil moisture holding capacity, crop water requirement may also affected this. Moreover, Occurrence of rainfall in time is very important for proper grown up of crops. The correlation between same drought indices with seasonal rainfall is also observed. The meteorological drought index indicates that in both dry and wet seasons it is significantly correlated with rainfall (for all the stations the value of r =1.000). But for agricultural drought index the degree of correlation vary from station to stations. And during dry seasons the strength of correlation is more significant than wet season. In dry season, degree of Pearson correlation vary from r=0.472 to r=0.413 but in wet season the degree of Pearson correlation vary from r=0.137 to r=0.801). It might be the cause that degree of rainfall variation is much higher in wet season. And this variability of rainfall negatively impacted the other component of agriculture that is closely related to the agricultural production. Moreover the occurrence of rainfall in this region is very low during dry season 519

10 Figure 8: Significant spatial correlation of meteorological and agricultural drought in Premonsoon season Figure 9: Significant spatial correlation of meteorological and agricultural drought in monsoon season 520

11 SPI maps indicate that meteorological drought in the Barind region appears randomly in the monsoon season. Irregular drought pattern in this region is due to inconsistent rainfall distribution (Figure 8) during monsoon. Meteorological drought scenario in the region changed continuously in monsoon season depending upon rainfall amount and spatial distribution. But during the pre-monsoon season the spatial distribution of meteorological drought showed a very similar pattern. The drought severity is found be higher in the northeastern part of the study area (Figure 8). Agricultural drought neither follows any spatiotemporal pattern like meteorological drought. The meteorological and agricultural drought indices show a very poor spatial correlation. During the dry season among 40 years only four years shows spatially significant correlation (figure 8). The years are 1975, 1990, 1998 and But the degree of correlation is also very significantly. The figure 8 evident that lower severity of meteorological drought can generate higher severity of agricultural drought. For example, in the year of 1975 a discrete pocket of northeastern Barind affected by chronic agricultural drought but almost entire part is affected by agricultural drought. Moreover, in the year 2003 few discrete pocket of southern part of the study area affected by mild meteorological drought but in the same time the entire southern part was affected by chronic agricultural drought. SPI and Markov chain maps indicate that a deficient rainfall as per the SPI index does not always correspond to agricultural drought. On the contrary, drought may appear in the agricultural spheres in spite of normal rainfall according to SPI. From the figure 9 it is clear that the severity of agricultural and meteorological drought also varied in the same time period. For example in the year 1982 the southern Barind affected by chronic to moderate meteorological drought but agriculturally severe to moderate drought has been affected the area. Figure 8 and 9 also evident that the lack of spatial correspondence of SPI and Markov chain is more prominent during the non-monsoon season. 5. Conclusion The SPI maps indicate that meteorological drought appears in Barind region in a random fashion in monsoon season but in pre monsoon it exhibits a similar pattern. However the seasonal meteorological drought has a short life span. The Markov chain maps reveal that during Pre-Kharif growing season the agricultural sphere of Barind region is seriously affected by chronic drought almost every year. But in Kharif season it demonstrated random fashion in some discrete pockets. The magnitude and frequency of agricultural drought vary from years to years in different locations. The study also identified that there is a very strong statistically significant correlation between SPI and Markov chain drought indices but they are poorly spatially correlated. In addition to this speed of drought development and drought duration vary widely. Therefore identification, classification and analysis of drought dynamics are influenced by monitoring parameter and monitoring methods. SPI monitor precipitation deficit the primary cause of drought development but a very little impact of agricultural drought have been observed through precipitation deficit. Despite a negative SPI a region can be free from drought or vice versa. Thus meteorological drought does not always correspond to agricultural drought. 6. References 1. Alam, A. T. M. J., Saadat, AHM., Rahman, M. S., and Barkotulla, M.A.B. (2011), Spatial analysis of rainfall distribution and its impact on agricultural drought at barind region, Bangladesh. Rajshahi University Journal of Environmental Science. 1(1), pp

12 2. Alam, A. T. M. J., Rahman, M. S., Saadat, AHM. (2012), Comparison of threshold values of Markov chain for determining agricultural drought at Barind, Bangladesh, North Bengal Drought Conference (NBDC) 2012 on Sharing Knowledge Combating Climate Change Disaster, University of Rajshahi and Kakonhat, March, pp Banik, P., Mandal, A., and Rahman, M. S. (2002), Markov chain analysis of weekly rainfall data in determining drought-proneness. Discrete Dynamics in Nature and Society, 7, pp Batisani, N. (2011), The Spatio-Temporal-Severity Dynamics of Drought in Botswana, Journal of Environmental Protection, 2(6), pp Bhuiyan, C., Singh, R.P., and Kogan, F.N. (2006), Monitoring drought dynamics in the Aravalli region (India) using different indices based on ground and remote sensing data. International Journal of Applied Earth Observation and Geoinformation 8, pp Biamah, E.K., Sterk, G., and Sharma, T.C. (2005), Analysis of agricultural drought in Iiuni eastern Kenya: application of markov model. Hydroligical Processes., 19, pp Christensen, J.H., Hewitson, B., Busuioc, A., Chen, A., Gao, X., Held, I., Jones, R., Kolli, R. K., Kwon, W.T., Laprise. R., Magan a Rueda, V., Mearns, L., Mene ndez, C. G., Ra isa nen, J., Rinke, A., Sarr, A., Whetton, P., (2007), Regional climate projections. In: Solomon SD, Qin M, Manning Z, Chen M, Marquis KB, Averyt MT, Miller HL (eds) Climate change 2007: the physical science basis. Contribution of working group I to the fourth assessment report of the intergovernmental panel on climate change.cambridge University Press, Cambridge 8. Darcup, J. A., Lee, K.S., and Paulson, E.G. (1980), On the definition of drought. Water Resources Research, 16, pp Dey, N. C., Alam, M. S., Sajjan, A. K., Bhuiyan, M A., Ghose, L., Ibaraki, Y.,and Karim, F. (2011), Assessing Environmental and Health Impact of Drought in the Northwest Bangladesh. Journal of Environmental Science & Natural Resources, 4(2), pp Goovaerts, P. (2000), Geostatistical approaches for incorporating elevation into the spatial interpolation of Rainfall, Journal of Hydrology, 228 (1), pp Heim, R.R. (2002), A review of twentieth-century drought indices used in the United States. Bulletin of the American Meteorological Society 83(8), pp Komuscu, A.U., (1999), Using the SPI to analyze spatial and temporal patterns of drought in Turkey. Drought Network News 11 (1), pp Loukas, A., and Vasiliades, L. (2004), Probabilistic analysis of drought spatiotemporal characteristics in Thessaly region, Greece, Natural Hazards and Earth System Sciences 4, pp Loukas, A., Vasiliades, L. and Tzabiras. J., (2008), Climate change effects on drought severity. Adv. Geosci., 17, pp

13 15. McKee, T.B., Doesken, N.J. and Kleist, J. (1993), The relationship of drought frequency and duration to time scales. 8th Conference on Applied Climatology, January, Anaheim, CA, pp McKee, T.B., Doesken, N.J., Kleist, J. (1995), Drought monitoring with multiple time scales. In: Proceedings of the 9th Conference on Applied Climatology, Am. Meteorol. Soc. Boston, pp Mishra, A.K., and Singh, V.P. (2010), A review of drought concepts. Journal of Hydrology 391, pp Mishra, A.K., Desai, V.R., Singh, V.P. (2007), Drought forecasting using a hybrid stochastic and neural network model. Journal of Hydrologic Engineering. 12 (6), pp Mishra, A.K., Singh, V.P., (2009), Analysis of drought severity-area-frequency curves using a general circulation model and scenario uncertainty. Journal of Geophysical Research. 114, pp Mishra, A.K., Singh, V.P., Desai, V.R., (2009), Drought characterization: a probabilistic approach. Stochastic Environmental Research and Risk Assessment. 23 (1), pp Moslehuddin, A.Z.M., Habibullah, Moniruzzaman, M. and Egashira, K., (2008), Mineralogy of soils from different agroecological regions of Bangladesh: region 25 level Barind tract and region 27- north eastern Barind tract, Journal of. Faculty of Agriculture Kyushu University. 53 (1), pp Mpelasoka, F., Hennessy, K., Jones R. and Bates, B. 2008, Comparison of suitable drought indices for climate change impacts assessment over Australia towards resource management. International. Journal of Climatology. 28: Ochola, W.O., Kerkidis, P. (2003), A Markov chain simulation model for predicting critical wet and dry spells in Kenya: analysing rainfall events in the Kanoplains, Irrigation and Drainage., 52, pp Paul, B.K. (1998), Coping mechanisms practiced by drought victims (1994/5) in North Bengal, Bangladesh. Applied Geography, 18, pp Quiring, S. M., (2009), Monitoring Drought: An Evaluation of meteorological drought indices, Geography Compass, 3(1), Rahman, M. S. (1999a), A stochastic simulated markov chain model for daily rainfall at Barind, Bangladesh. Journal of Interdisciplinary Mathematics, 2(1), pp Rahman, M. S. (1999b), Logistic regression estimation of a simulated Markov chain model for daily rainfall in Bangladesh. Journal of Interdisciplinary Mathematics. 2(1), pp Rahman, M. S. (2000), A rainfall simulation model for agricultural development in Bangladesh. Discrete Dynamics in Nature and Society, 5, pp

14 29. Saadat,A.H.M., Alam, A.T.M. J., Alam. M., Shovon,J. and Uzzaman, R. (2009), Impact of drought on agriculture of Barind tract: A case study of Dharmapur Chapainawabgang. Workshop on Impacts of Climate Change on Livelihoods, Agriculture, Aquaculture and Fisheries Sector of Bangladesh, BAU, Mymensingh, October 1, 2009, Ramsey, S., Subbia, A. R., Bass, S. and Juergens, I. (2007), Livelihood adaptation to climate variability and change in drought- prone areas of Bangladesh, Asian Disaster Preparedness Center Food and Agriculture Organization of the United Nations. 31. Shahid, S and Behrawan, H, (2008), Drought risk assessment in the western part of Bangladesh. Natural Hazards, 46, pp Shahid, S and Hazarika, M. K., (2010), Groundwater Drought in the Northwestern Districts of Bangladesh, Water Resource Management, 24, pp Shahid, S, (2008) Spatial and temporal characteristics of droughts in the western part of Bangladesh. Hydrological Processes. 22, pp Shahid, S, (2010) Rainfall variability and the trends of wet and dry periods in Bangladesh, International Journal of climatology, 30, pp Shahid, S. (2011), Impact of climate change on irrigation water demand of dry season Boro rice in northwest Bangladesh. Climatic Change, 105, pp Sharma, T. C. (1996), Simulation of the Kenyan longest dry and wet spells and the longest rain-sums using a Markov model, Journal of Hydrology, 178, pp Steinemann AC, Hayes MJ, Cavalcanti LFN. (2005), Drought indicators and triggers. In Drought and Water Crises: Science, Technology, and Management Issues, Wilhite DA (ed). CRC Press: Boca Raton, pp Wilhite, D. A. and Glantz, M. H. (1985), Understanding the drought phenomenon: The role of definitions. Water International, 10(3), pp