Study on land-use and land cover change of Dhaka eastern region and its impact on water cycle

Abstract

ABSTRACT This study was focused mainly to identify the impending land-use and land-cover (LULC) change of Dhaka Eastern Region (DER) due to urbanization and assess its impact on two important hydrological parameters of water cycle, i.e., surface runoff and groundwater (GW) recharge. Existing LULC changes of DER were mapped from Landsat 5 (TM), 7 (ETM+) and 8 (OLI-TIRS) imageries respectively for 2010, 2012 and 2016-17 using supervised classification algorithms. The classified Landsat imagery of 2016-17 detected the direction of likely post structural urban development envisaged in Regional Development Plan (RDP) for DER. A slower pace of urbanization was traced in terms of paved built-up coverage with 4% in 2010, 6% in 2012 and 8% in 2016-17 from the total 215.28 km2 land of DER. Conversely, a faster pace from 8% in 2010 to 20% in 2016-17 was observed in case of unpaved bare soil or filled land transformed mainly from agricultural and tree-covered areas which might be built-up paved surface with a likely 65% post-developed imperviousness by 2035 or beyond as projected in RDP scenario. RDP proposed LULC had significant increase in paved built-up areas from existing 8% to 49% by 2035. Besides, four post-development LULC changes were predicted from the classified Landsat 8 imagery where predicted seneario-1 followed RDP scenario with nearly 62% paved surface mainly due to conversion of identified unpaved tree-covered area as opposed to conversion of agricultural areas in case of scenario-2. However, predicted scenario-3 and 4 with maximum urbanization resulted nearly 94% paved surface with respectively 65% and 85% imperviousness. This study estimated runoff by SCS-CN (Soil Conservation Service- Curve Number) method and recharge by WTF (Water-table Fluctuation) and four empirical methods. Like the slower trend of urbanization, CN also increased slowly from 78 in 2010 to 81 in 2016-17, but followed a moderate leap in RDP and scenario-1 with CN 84. However, CN dropped to 78 in scenario-2 mainly due to change in land-cover from low-permeable agricultural soils to highly permeable sandy soils. Predicted scenario -3 and 4 resulted CN 82 and 90 respectively. In fact, higher CN resulted higher runoff and more recharge loss. Hydrological impacts on runoff and recharge were assessed from the pre-developed 2010 LULC as a reference year. Two different sets of impact on runoff were observed due to average and seasonal variation of antecedent moisture condition (AMC) of soils. Distinctively, seasonal AMC generated around 14% to 17% higher runoff than average AMC. Similar to runoff, both seasonal and average AMC had varied influence on GW recharge. Estimated pre-developed natural and net recharge potentiality of DER with 337 mm/year and 488 mm/year respectively declined gradually due to LULC change except for predicted scenario-2. Having assessed the overall impacts of LULC change on runoff and recharge, scenario-2 is hydrologically more suitable than others. But, the direction of LULC change indicates RDP or predicted scenario-1 is an impending LULC of DER. In effect, RDP/scenario-1 (CN 84) was considered as the limiting LULC for DER which would approximately result 40% runoff and 50% reduced natural recharge annually with seasonal AMC. However, predicted scenario-4 with maximum CN 90 would result more than 50% runoff and complete cut-off of both natural and net recharge potentiality with a significant decline in GW level which indeed, would depict the typical landscape of core city Dhaka within DER.