HYDROLOGICAL RESPONSE SIMULATION UNDER LAND USE AND CLIMATE CHANGE SCENARIOS

Abstract

Water is one of the most essential components of the environment and requires proper planning and management to achieve its sustainable utilization. Changes in climate and land use have significantly influenced the hydrological cycle and hence affect water resources. Due to uncertainty in climate change projection and land use change, improved knowledge of basin hydrology and resources availability are indispensable for policy formulation and sustainable development of the water sector. The aim of this research was to improve understanding of hydrological response for a large river basin under land use and climate change scenarios. This study investigates the changes in water balance components under different land use and climate change scenarios over Upper Narmada Basin (UNB), a river basin in the sub-humid area of central India. Moreover, identification of climate change and selection of climate models were carried out at regional scale as well as local scale (UNB). For regional scale studies, India has been categorized in to seven zones, considering the geography and homogeneous annual precipitation i.e. North Mountainous India (NMI), North Central India (NCI), Northwest India (NWI), East Peninsular India (EPI), West Peninsular India (WPI), South Peninsular India (SPI) and North East India (NEI). The comprehensive assessment of spatial and temporal variability in annual and seasonal precipitation and temperature was carried out at regional scale and local scale. In order to identify climatic variability over the region, parametric and non-parametric tests were performed to detect trend, periodicity and break points in long term precipitation and temperature data. At regional scale (seven Indian zones), regression analysis was carried out on long term (1851-2006) monthly precipitation. The results imply mean values of precipitation are decreasing in most of the zones in the last 30-year period while both positive and negative trends existing in each zone for the monsoon datasets. Therefore, wavelet analysis is quite popular tool for trend analysis and periodicity identification in hydrological time series. Discrete Wavelet Transform (DWT) Daubechies wavelets db6 and db10 were selected to decompose the annual and monthly datasets, respectively, applying criteria minimum MRE and the minimum criterion relative error (Er). The Z statistic was evaluated for trend analysis of the decomposed periodic components and the original annual and monsoon series. Application of DWT on annual series implied 2-, 4- and 8-year fluctuations in the NMI zone, indicating a positive trend in rainfall, whereas zones WPI, SPI and WPI (with 2- and 4-year fluctuations) vi | P a g e experienced a negative trend at the same periodicities, at the 0.05 (5%) significance level. In the monsoon series, a positive trend was found over NMI and NEI decomposed series at 2-, 4- and 8- year periodicity, whereas WPI, EPI and SPI indicated a negative trend at the same periodicity. Considering India as whole (AI), it was found a negative trend in all zones except NMI and NEI. Furthermore, at local scale analysis, trend in monthly and annual precipitation and mean temperature data were identified at 16 stations of UNB. However, results imply that 8% stations exhibit the negative trend for precipitation, while 100% stations show positive trend for mean temperature over 16 stations of UNB. A comparative study was carried out between three methods i.e. innovative trend analysis (ITA), Mann Kendall (MK) test and Sen’s slope, to check the suitability of ITA against nonparametric tests. In result, ITA show the strong agreement with both methods (MK test and Sen’s slope), 97.5% and 77.5% in ‘ITA versus MK test’ and ‘ITA versus Sen’s slope’. Change year obtained from sequential Mann Kendall (SQMK) were compared with the change year of CUMSUM. Results indicated most of the stations exhibit significant abrupt change year is 1955 (77.78%) for precipitation, and 1960 (100%) for mean temperature. However, finding from this analysis has improved understanding of variability, at spatially and temporally, at current and future climate change. Evaluation of changes in land use land cover are extremely important and must be monitored to assess the impact on environment. In this study, mapping of LULC and change detection were carried out using the Landsat TM satellite image and geospatial tools. The development of LULC classes were evaluated from 1990 to 2000 to 2010 to 2015. The reduction in natural vegetation and increase in settlement as well as cropland are reflected in the analysis of LULC mapping. Understanding of trend patterns were demonstrated and predicated for the year 2030 using CA-Markov model. The model were validated with simulated and actual LULC of 2015. The projected LULC of 2030 classes indicated the continuing of same trend of recent past. These future projection indicate the expected changes in near future. Therefore, the LULC changes in classes in near future recommend the planning and management of the study area. To assess the impact of climate change on hydrology, Global Climate Models (GCMs) are the primary data source. Therefore, performance of GCM models are required to evaluate for choosing the best representative GCM for the region. To identify the best representative climate models, performance of 24 GCMs were evaluated against reanalysis models, based on Skill Score (SS) and Root Mean Square Error (RMSE) of six climatic variables. After applying multi-criteria analysis on evaluation parameters (SS and RMSE), results indicate that there is no single vii | P a g e GCM which can recommend for whole Indian regions. However, GCM models for representation of precipitation have been proposed as ensemble of MPI_ECHAM4.0, MIROC3.2_HIRES, UKMO_HADCM3.0 and INGV_ECHAM4 for Indian regions. Climate models of CCCMA groups are performing well for atmospheric temperature in most of the Indian regions. At local scale, three climate models MIROC5, CNRM-CM5 and MPI-ESR-LM were evaluated as best performing models for hydrological modelling of future climate change over UNB. In order to assess the hydrological response under land use and climate change, Soil and Water Assessment Tool (SWAT), a semi-distributed hydrological model was calibrated applying multi-site calibration techniques. In monthly simulation, Nash Sutcliffe Efficiency (NSE) and Coefficient of Determination (R2) were computed, 0.77 and 0.76 for calibration (1978-1995) and 0.73 and 0.70 for validation period (1996-2005), respectively, indicating good performance for basin. Calibrated hydrological model used to simulate the change in water balance components under three different land-use and two climate-change scenarios from three representative GCMs (MIROC5, CNRM-CM5 and MPI-ESR-LM). Hydro-meteorological response under land use change indicate that increase in settlement and decrease in natural vegetation, affect as increase in the water yield and surface runoff, but decrease in evapotranspiration (ET). The actual ET decreases with time due to decrease in natural vegetation, maximum in 1990 (460.04 mm) while projected in 2030 is lower (407.19 mm). Water balance components under climate change scenarios indicated annual precipitation decreasing from -1.65% (MPI) to -16.55 % (MIROC) during P1 (2011-2040) and P2 (2041-2071) under RCP4.5, whereas in RCP8.5 scenarios it varies from -26.19% (CNRM) in P3 (2071-2100) to 21.24 % (MIROC) in P3, with reference to baseline scenario. Changes in green and blue water varying from 16.22% (MIROC, P3) to -14.10% (CNRM, P3) under RCP4.5 and from 38.25%(MIROC, P3) to -22.57% (CNRM, P3) under RCP8.5 with reference to baseline scenario. This study established the sensitivity of UNB to future climatic changes employing projections from CMIP5 climate models and exhibited an approach that applied multiple climate model outputs to estimate potential change over the river basin. Moreover, adaptation strategies are proposed against climate change impact. In general, the study provide a scientifically important and practically relevant, to identifying the historical climate variability and hydrological assessment under land use and climate change scenarios considering representative climate models output, in contributing to water resources planning and management in the context of river basin.