From drought deficits to climate resilience: coupled surface-groundwater modelling and nature-based solutions for drought
|  |
| Available in | Authors |
|
Document type: Dissertation
|
| Project | Top | Authors |
- PhD - Predicting Impact Climate Change on Drought in the Scheldt River Basin, more
|
| Authors | | Top |
- Yimer, E.A.
- Van Griensven, A., revisor, more
- Nossent, J., revisor, more
- Huysmans, M., revisor, more
|
- Tysmans, T., revisor
- De Graeve, I., revisor, more
- Van Loon, A., revisor
- Mustafa, S.M.T, revisor
|
|
| Abstract |
Geohydrological modelling is crucial for understanding how water balance components react to changing environments. Researchers usually prefer to model surface and groundwater processes separately due to the difficulty of coupling those water systems. In doing so, groundwater-surface water interactions will be ill-represented. Thus, coupled models are required mainly in groundwater-dominated watersheds with shallow aquifers where the streams and aquifers interact constantly. The widely applied soil water assessment tool (SWAT) was recently modified with versatile options representing detailed processes (SWAT+). Despite improvements, groundwater representation is still conceptual. Thus, recent advancements have introduced a groundwater flow module (gwflow) for SWAT+, making it fully physically based and distributed. In this thesis, the newly coupled SWAT +gwflow model is modified to account for groundwater-soil interaction and produced aiding tools for model development, calibration and post-processing. The enhanced model is then tested and validated in the Scheldt basin - specifically the Kleine Nete, Grote Nete, Demer, Dijle and Zenne, Dender, Upper Scheldt, and the Leie as well as to some selected representative watersheds across the globe.
Models are vital to evaluate the extent to which climatic extremes affect surface and groundwater reserves. As hydroclimatic extremes are becoming more frequent and intense, investigating how these events evolve spatially and temporally is crucial. Specifically, drought has complex patterns and occurring mechanisms affecting several layers of watershed water balance components. Thus, in this research, the SWAT+gwflow model setup for the Scheldt basin is used to evaluate the impact of climate change on streamflow and groundwater droughts. There are two major drought analysis methodologies, one which is a standardized index based which fits a given variable (e.g., precipitation, streamflow, groundwater head, etc.) to a probability distribution function and then transforming it to cumulative distribution function and ultimately converting it to standardized normal distribution function making it spatially and temporally comparable. This method fails to provide information on the amount of deficit. As policymakers are keen to know how much the decline in a given water balance component is due to drought, threshold-based drought analysis is the widely used method. Therefore, this research compares the two methods, generates a methodology to find an appropriate distribution function, and ultimately uses an attribution analysis using the Inter-Sectoral Impact Model Intercomparison Project (ISIMIP) factual (with climate change) and counterfactual (without climate change - detrended data) data to evaluate and assess if climate change is a key player and to what extent it aggravates hydrological drought events.
In recent years, nature-based solutions (NbS) for hydroclimatic extreme mitigation/ adaptation measures have been gaining momentum. These measures opt for natural means of increasing the landscape’s water holding capacity, allowing groundwater systems to recharge, decreasing runoff, reducing sediment load, etc. Drought is the least studied hazard compared to floods, regarding NbS as a mitigation measure. Also, there is a lack of framework for suitability mapping of NbS locations for drought mitigation. Therefore, in this last work package, an extensive literature review is made to assess NbS targeting drought, henceforth, a new framework for finding locations with the highest possible advantage on recharging aquifers is investigated.
The results indicated that the standalone SWAT+ model is unable to capture the geohydrological water balance components and river discharge (Nash-Sutcliffe efficiency (NSE)
below 0.5). However, the coupled model with the inclusion of groundwater-soil interaction has improved geohydrological simulations by providing reasonable annual average water balance and monthly streamflow values. Next, the regional groundwater-surface modelling of the Scheldt basin indicated successful application of the coupled model with calibration and validation of streamflow, showing NSE ranging between 0.6 and 0.9. The seven watersheds inside the regional Scheldt basin mostly have shallow aquifers, making the modelling practice convenient with the SWAT+gwflow model setups. Furthermore, an investigation of the Kleine Nete watershed to simulate agricultural water drainage flux has added to the coupled model’s ability to capture complex geohydrological conditions. In the agricultural water drainage assessment, finding the locations of ditches was critical, hence, a novel methodology was developed that considers groundwater levels along with classical remote sensing and decision classification methods. Finally, the development of an open-source Python script for the coupled model input preparation (model development), calibration, and post-processing eased the coupled model’s application among the wider SWAT community.
Next, by using the calibrated and validated coupled models for the seven watersheds, an attribution study was made to evaluate the extent of climate change impact on rivers and groundwater and hydrological drought in general. Factual and counterfactual data from the Inter-Sectoral Impact Model Intercomparison Project (ISIMIP) is used to force the models. The impact assessment on the rivers has indicated that the low flows decreased between 3.5 and 14%. The cumulative deficit showed the highest impact on the Upper Scheldt catchment. Furthermore, the Upper Scheldt and Leie basins experienced an average increase in streamflow deficit of about 10 m3/s during drought periods. The use of three detrended datasets has indicated uncertainty of deficit values, but all agree that the impact of climate change on the rivers is evident. The groundwater attribution on the Kleine Nete indicated that the groundwater level has decreased considerably in recent years (2010 onwards). Overall, from 1970 onwards, there was an average of a 10 cm significant decline in groundwater head, with some locations having a maximum decline ranging between 0.1 and 1.5 meters, which shows the uncertainty spread. The groundwater drought assessment has indicated an increase in maximum drought deficit and duration due to climate change. The deficit is mainly high for short-term (1 to 3 months) and yearly impact but revives at a 24-month timescale. This is attributed to the shallow nature of the aquifer water table, which declines in shorter time and revives at longer timescales. These highlight the need to apply necessary mitigation or adaptation measures to reduce the impact climate change causes on the Kleine Nete or, in general, unconfined groundwater reserves.
After highlighting that drought is partly attributed to climate change and the need for mitigation/adaptation measures is found to be evident, NbS mainly targeting drought are investigated. A literature review made on NbS and drought indicated the following major findings: few NbS studies focus on drought compared to other hazards such as floods,
most NbS implementations are located in Europe, effectiveness is not well quantified for implemented NbS and finally, for achieving considerable drought mitigation, large-scale NbS are needed. With this in mind, the next research focuses on finding large-scale NbS, which requires framework development, and applying the framework at a regional scale, with Flanders as a case study area. The main findings from this work were the following: The new mapping framework addresses drought-specific needs and climate scenarios, and the results for Flanders underscore the highly suitable potential for detention basins (2552.2 km²) and managed aquifer recharge areas (2538.7 km²), emphasizing the adaptability and scalability of the framework for addressing drought in the region, detention basins compensate partly for the high groundwater demand, and finally, regional application yields actionable insights for decision-makers.
This dissertation emphasizes that the physical and distributed nature and modification we applied to the gwflow module have enhanced geohydrological simulations in groundwater-dominated watersheds, such as the Scheldt basin. Additionally, climate change impacts on surface and groundwater are analyzed by estimating deficit values that policymakers can use to assess extremes and their possible impact. Finally, resilience against drought with NbS is assessed by locating potential NbS areas based on several conditions. Thus, water managers can utilize this research to apply necessary mitigation measures to tackle the hazard by looking into its impact at several layers of the Scheldt River basin.
This PhD dissertation is structured around four major parts. The first focuses on model development, including model modification, verification, and tool development. Secondly, using the tools and modifications made in the first part, the model is further used for assessing processes such as drainage impact on geohydrology, groundwater–surface water interactions and groundwater–soil interactions in the Scheldt basin (model application). Thirdly, the calibrated and validated model is used to investigate drought in the region. Finally, solutions for drought mitigation are explored, focusing mainly on nature-based |
|
