Estimating the Effects of Climate Change on the Water Resources in the Upper Indus Basin (UIB)
Water balance calculations and spatially distributed rainfall-runoff models require high-resolution climatic datasets as the primary input, however, in the Upper Indus basin (UIB), it is rare to find in-situ observational climate data which have spatial, altitudinal and temporal coverage, suitable enough for distributed hydrological investigations and so for the assessment of climate-change’s hydrological impacts. The average precipitation amounts over the UIB are unrealistically low to sustain the observed discharge at the basin outlet. This is mainly due to the presence of sparse observational networks, mostly comprised of valley-based, low-altitude gauge stations, incapable to capture the orographic effects in the mountainous UIB. The solution to these issues may be either the use of gridded data products or the generation of improved data sets utilizing state of the art of science and techniques. The first four (4) chapters of the present thesis are dedicated to address the issues of the poor quality of hydroclimate data in the UIB. The remaining two chapters (5 and 6) address the main aim of the study, i.e. the projections of future hydrological scenarios, and the preceding necessary activities, including calibration and validation of the SWAT hydrological model, as well as the selection, downscaling and bias correction of climate models’ projections. A brief account of each chapter is given in the following paragraphs: 1. This part of the present study aims to evaluate the capability of the “Tropical Rainfall Measurement Mission” (TRMM) “Multi-satellite Precipitation Analysis” (TMPA) to estimate appropriate precipitation rates in the Upper Indus basin (UIB) and to analyse the dependency of the estimates’ accuracies on the time scale. To that avail statistical analyses and comparison of the TMPA- products with gauge measurements in the UIB are carried out. The dependency of the TMPA estimates’ quality on the time scale is analysed by comparisons of monthly, seasonal and annual sums for the UIB. The results show considerable biases in the TMPA- (TRMM) precipitation estimates for the UIB, as well as high false alarms and miss ratios. The size of the correlation of the TMPA- estimates with ground-based gauge data increases considerably and almost in a linear fashion with increasing temporal aggregation, i.e. time scale. The BIAS is mostly positive for the summer season, while for the winter season it is predominantly negative, thereby showing a slight over-estimation of the precipitation in summer and under-estimation in winter. The results of the study suggest that, in spite of these discrepancies between TMPA- estimates and gauge data, the use of the former in hydrological watershed modelling, endeavoured in the subsequent chapters, may be a valuable alternative in data- scarce regions, like the UIB, but still must be taken with a grain of salt. 2. Six interpolation methods are employed in this study to improve the spatial interpolation of precipitation data in the Upper Indus Basin (UIB), including NN, IDW, OK), SK, KED and SKlm. Quantitative cross validation shows that out of these methods, SKlm performs better for data aggregates at monthly, seasonal and annual time scales, according to the values of several scoring indices, including r, MAE, RMSE, and NSE. The performance of the SKlm method is followed in decreasing order by SK, OK, KED, IDW and NN. According to the results of a qualitative cross validation analysis, based on six categorical indices for the daily interpolation estimates (Ac, FBI, POD, FAR, CSI and TSS), the estimates generated by SKlm, with an average rank of 1.83, are also better than those of the methods tested. The remaining methods follow the similar performance pattern as indicated by the quantitative indices. In conclusion, SKlm proves to be overall the best option for interpolating precipitation data in the UIB, by providing a comparatively better representation of the latter, both in terms of magnitudes as well as occurrences. 3. In this section of the current study a new approach for the interpolation and regionalization of observed precipitation series to a smaller spatial grind scale (0.125° by 0.125°) across the UIB, with appropriate adjustments for the orographic effect and changes in glacier storage , is evaluated and validated through reverse hydrology, guided by observed flows and available knowledge base. More specifically, the generated corrected precipitation data is validated by means of SWAT-modelled responses of the observed flows to the different input precipitation series (original and corrected ones). The results show that the SWAT- simulated flows using the corrected, regionalized precipitation series as input are much more in line with the observed flows than those using the uncorrected observed precipitation input for which significant underestimations are obtained. 4. The methodology used in the previous Chapter-3 is combined here with a few other methods to address another major issue prevalent in the UIB which is the non-existence of long-term climate records. To resolve this issue, precipitation time series are generated at virtual locations through temporal reconstruction of the precipitation series at those points where data was not recorded prior to the mid-nineties. The data generated at these virtual locations is corrected with the 12-years observed mean there, followed by regionalization of the precipitation series (Chapter-3) to a smaller scale across the basin. The precipitation data reconstructed in this way is validated again through evaluations of their SWAT-modelled streamflow responses. The results show that the daily simulated discharges obtained with the calibrated SWAT- model forced with the reconstructed precipitation are almost identical to those acquired when forcing the model with the reference precipitation. 5. This section of the study focusses on identifying a set of representative future climate projections for the UIB. Although a large number of GCM’s predictor sets are nowadays available in the CMIP5 archive, the issue of their reliability for specific regions must still be confronted. This situation along with other factors such as time, human resources or computational constraints, makes it imperative to sort out the most appropriate, single or small-ensemble set of GCMs for the assessment of climate change impacts in a region. Here a various approaches are adopted and applied for a step-wise shortlist and selection of appropriate climate models for the UIB for two representative concentration pathways (RCPs), RCP 4.5 and RCP 8.5, based on, a) range of projected mean changes, b) range of projected extreme changes, and c) skill in reproducing the past climate. Furthermore, because of higher uncertainties in climate projection for high mountainous regions, like the UIB, in addition to projections that provide the closest representation of the mean future climate change, all possible future extreme scenarios (wet-warm, wet-cold, dry-warm, dry-cold) are also considered. Based on this two-fold procedure a limited number of climate models is pre-selected, out of which the final selection is done by assigning ranks to the weighted score for each of the mentioned selection criteria. The dynamically downscaled climate projections from the Coordinated Regional Downscaling Experiment (CORDEX) available for the top-ranked GCMs are further statistically downscaled (bias-corrected) over the UIB, for subsequent use as climate drivers in the SWAT- model. 6. Projecting future hydrology for the mountainous, highly glaciated upper Indus basin (UIB) is a challenging task, because of uncertainties in the future climate projections and issues with the coverage and quality of available reference climatic data and hydrological modelling approaches. This part of the study attempts to address these issues by utilizing the semi-distributed hydrological model ‘SWAT’ with new climate datasets with better spatial and altitudinal representation (Chapter-4) as well as a wider range of future climate forcing from the CORDEX GCM/RCM-model set (Chapter-5), to assess different aspects of the future hydrology (mean flows, extremes and seasonal changes). Contour maps for the mean annual flow and actual evapotranspiration as a function of downscaled projected mean annual precipitation and temperatures are produced which can serve as a simple “hands-on” forecast tool of the future hydrology in the UIB. The overall results of these future hydrological projections show that almost all RCP- climate scenario/ GCM/RCM- model combinations lead to similar trends of changes in magnitudes, seasonal patterns and extremes of the UIB streamflows. In particular, all, but one scenario - that with a very high future temperature rise - indicate increases of the mean annual flow throughout the 21st century, wherefore, interestingly, these are stronger for the middle of the century (2041-2070) than at its end (2071-2100). The seasonal shifts as well as the extremes follow the same trend for all climate scenarios, such that an earlier arrival (in May-June instead of July-August) of high flows and increased spring and winter flows, whereas the upper flow extremes (peaks) are projected to drastically to increase by 50 to >100%, and this with decreased annual recurrence intervals, i.e. there will be a tremendously increased future flood hazard in the UIB. The future low flows projections also show more extreme values, with lower than nowadays-experienced minimal flows, occurring more frequently and with much longer annual total duration, as well, most likely due to increased future glacier melting.
@phdthesis{doi:10.17170/kobra-2018120951, author ={Khan, Asim Jahangir}, title ={Estimating the Effects of Climate Change on the Water Resources in the Upper Indus Basin (UIB)}, keywords ={620 and Klimaänderung and Wasserreserve and Wasserbilanz and Hydrologische Vorhersage and Fehlende Daten}, copyright ={https://rightsstatements.org/page/InC/1.0/}, language ={en}, school={Kassel, Universität Kassel, Fachbereich Bauingenieur- und Umweltingenieurwesen}, year ={2018} }