Predicting streamflow in a changing climate poses significant challenges for traditional hydrological models. Static parameter sets result from model calibrations over historical data that increasingly encounter the non-stationary impacts on the hydrological system. Endeavouring toward non-stationary model parameters by incorporating time-adaptive ecosystem-scale root zone storage parameters shows promise for modelling systems under change. The ongoing CATAPUC project aims to further develop, refine, and implement this adaptive modelling approach. This paper continues to build upon this body of work by investigating the evidence for land cover change impacts on root zone storage capacity, focusing specifically on uncertainties in the evaporative balance. Combining long-term hydrometeorological data from the primary study area, the Meuse basin, with the large-sample CAMELS datasets (GB and US) analyses were performed across 283 catchments. Applying the vector operations from Velde et al. (2014) and Jaramillo et al. (2018) to decadal changes within the Budyko framework, we separate climate-related evaporative changes. We isolated the residual component of evaporative change, which is the unknown component affecting parameter estimations. Our findings indicate that this residual component is twice as prominent in evaporative change as the climate component. The aim is to test if land cover changes have contributed significantly to the residual component of evaporative change. A multi-scale approach to land cover change analysis is adopted to bridge the data gap from 1984 to 2019. Implementing ensemble machine-learning methods on Landsat imagery with Google Earth Engine, we develop annual timeseries of (30m) high-resolution multi-class data for this period with accuracies up to 86% for 117 catchments (Meuse and GB). Resulting land cover change estimates suggest that Meuse Basin urbanisation rates may have been significantly underestimated for this period. The Meuse Basin over the most recent 20-year period is found to deviate anomalously in actual evaporation compared with the large sample. A distinct spatial pattern reveals a concentration of deviations in the east of the basin. Calculated by Tempel (2023), the anomaly in the basin corresponds with a relative median error in root zone storage capacity change of −14%. Low flow analysis is performed to remove the possibility that deviations are affected by anomalous contribution to streamflow from additional subsurface flow. We observed that the increase in low flow variability over the same period exhibits a spatial pattern similar to the Meuse Basin anomaly. We found no meaningful causal relationships linking multi-scale interdecadal changes in forest, agriculture, and urban land classes to the observed deviations in the large sample datasets. The implication of this study is that land cover change is likely not a significant driver of evaporative changes, specifically, errors, throughout the record available.

Climate change-induced changes in weather patterns call for the development of hydrological models that perform well under increasingly extreme and varied conditions. Multiple research studies have demonstrated that hydrological models perform poorly when applied to climate conditions that differ from those during the calibration period of the model (Duethmann, Bloschl, & Parajka, 2020). The need for robust hydrological models was emphasised after the Geul catchment, and large parts of Belgium and Germany flooded in July 2021. During this flood, the hydrologists from Waterboard Limburg (WL) indicated that the current model would not have been able to correctly forecast the flood under such an extreme rain event anyway (Expertise Netwerk Waterveiligheid, 2021).

The underlying assumption in this research is that hydrological models cannot perform to the same standard during changing weather conditions because they are overfitted during the calibration period. Meaning that the parameter values are the mathematical best fit for streamflow predictions, but do not represent the internal hydrological processes between precipitation and discharge. One way of trying to improve the internal processes of a hydrological model is to create more relations between these processes and external measurements (Kirchner, 2006). In this research, a secondary calibration data set is used, namely evaporation data.

There are two goals in this thesis, first building a hydrological model of the Geul which represents the streamflow response of the catchment in a meaningful way. Second, comparing if a difference in calibration methodology, specifically comparing calibration on discharge with calibration on both discharge and evaporation, would improve the predictive power of the model. These goals are set up to be able to answer the main research question of this thesis:

Is the predictive power of a discharge calibrated hydrological model of the Geul catchment in the Netherlands, greater than the predictive power of an evaporation calibrated model, in addition to discharge?

The methodology used in this thesis to calibrate and evaluate is Generalised Likelihood Uncertainty Estimation (GLUE), with NSE efficiency as objective function.

The first goal, building a hydrological model of the Geul which represents the stream flow response of the catchment satisfactory, is successfully reached. The model calibrated on streamflow (Model Q) and the model calibrated on both streamflow and evaporation (Model QE) were able to predict streamflow satisfactorily, with DeQ scores between 0.632 and 0.649 across all runs.

The second goal was to compare the performance of Model Q and Model QE. Model QE outperformed Model Q on monthly runoff coefficient, monthly average evaporation and cumulative evaporation, increasing the NSE score of Model Q of the monthly runoff coefficient from 0.774 to 0.801, the NSE log score of the monthly average evaporation from 0.884 to 0.891 and the cumulative evaporation from 0.760 to 0.774. However, Model QE did not outperform Model Q on streamflow.

Therefore, it cannot be concluded that additional calibration on evaporation data increases the predictive power for the streamflow of the model. However, the model now more accurately represents the observed evaporation data, without trading this in for less predictive power of streamflow. Therefore, Model QE can be assumed to be a more accurate description of the hydrological processes in the Geul catchment than Model Q.

Hydrological models are commonly used to predict future streamflow. However, the assumption of stationary model parameters obtained through calibration on past conditions may not accurately represent non-stationarity in hydrological system characteristics. Evidence suggests that vegetation adapts its root zone storage capacity in response to changing climate, emphasizing the need to account for non-stationarity in hydrological models. This study examines the effects of long-term climate variability on root zone storage capacity and its consequences on hydrological response. By using the method of Fu (1981), we determine whether the root zone storage capacity has significantly changed and evaluate the sensitivity of hydrological model predictions to these changes. We explain the changes in root zone storage capacity and evaporation using various climate indicators. For two large sample datasets CAMELS GB and CAMELS USA, we confirm that the Fu method can be used with the same omega parameter when transitioning from one decade to the next, indicating small differences in root zone storage capacity. However, for hydrometeorological data from the Meuse basin in Northwest Europe, we observe a trend where the actual evaporation is smaller than expected. This suggests that caution should be exercised when applying the Fu-method. In four different scenarios, we have implemented historical changes in evaporation as altered root zone storage capacity in the wflow flextopo model. The scenarios based on the evaporation trends observed in the Meuse data, result in less evaporation during the summer months (May, June, July) within a range of 0 to -22%, and an increase in streamflow during the autumn months (September, October, November) between 0 and +48%. On the other hand, the scenarios based on evaporation trends observed in all combined data (Meuse, CAMELS GB, CAMELS USA) result in changes around zero of evaporation during summer months between -7% and +5%, and of streamflow in autumn months between -11% and +10%. This study represents a step towards a more reliable and robust estimation of root zone storage capacity in hydrological modelling, thereby enhancing our ability to predict future streamflow.

In mid-July 2021, heavy precipitation caused flooding for multiple tributaries of the Meuse and Rhine rivers which resulted in many deaths and exceptional damage. The Ahr in Germany, the Vesdre in Belgium and the Geul in the Netherlands suffered the greatest damage in each country. The first step in reducing such high damage from international, minor catchments is to build an understanding of the flood event.

The goal of this study was to analyse how extreme several hydrological aspects of the flood event of July 13-16 were for the Ahr, Vesdre and Geul. Analysing the extremity of an event requires a description of that event which helps with building knowledge. Its extremity places that knowledge into context. A flood with a 10-year return period will require a different approach than one with a 1000-year return period to balance the risk reduction and required effort. Three international catchments have been selected as case studies to allow an in-depth analysis of catchments and compliment current research on the flood event with a transboundary perspective.

The hydrological aspects considered for this study concern the forcing, flow mechanisms and response. For the forcing and response, a data analysis has been performed to study the extremity of the origin, magnitude, spatial and temporal variation of the precipitation, as well as the magnitude of the flow. Characteristics of both were used to determine the flood type. The data analysis on flow magnitude and seasonality has been supplemented by historical information to give more insight. The flow mechanisms have been derived from two models with varying structure and spatial variation by analysing parameter values and internal states.

The origin of the precipitation is considered as extreme since the cold core low, a low pressure area with a cold centre, was accompanied by several factors in a rare combination that likely contributed to the high precipitation amounts. The precipitation amount of July 14, the wettest day of the event, was of extreme magnitude according to the return periods of over 1000 years for several gauges. Although there was a large spatial variation in precipitation and thus their return periods, the spatial pattern showed a positive correlation of medium strength with the climatic pattern and is thus not considered as extreme. The temporal variation of the event is also not extreme. So during the event, only the amount of precipitation was extreme, but the precipitation amount in the preceding month was also extreme.

The peak discharges have been estimated to be around 990, 600 and 90 $[m^3/s]$ at the downstream gauges of the Ahr, Vesdre and Geul. The higher peak discharges coincide with sharper hydrographs and a stronger relief and impermeable soils. The return periods of the peak discharges were over 1000 years. Such extreme return periods are not surprising as such discharges had never been measured before. However, historical reconstructions show that similar flood events have occurred at least once for the Ahr. The assumed rare occurrence of summer floods in the regions was also debunked by historical information as 21 to 44\% of the historical floods occurred during the summer half-year. Another assumption by reports on the flood event was the flood type identification of a flash flood. The large spatial and temporal scale, as well as lag times for downstream gauges of 13 to 24 hours, do not resemble a flash flood, but rather a long precipitation event with a fast response.

Two models, a lumped, simple model based on the HBV model and a distributed more complex model based on the SBM model, have been calibrated for several years preceding the flood event, resulting in a good performance. Their performance for the flood event was low due to a strong underestimation of the peak discharge. This underestimation was resolved by a re-calibration. The re-calibrated parameters and internal states of the models show an increase in fast responding mechanisms, up to unprecedented magnitudes. The SBM model includes more processes and shows a rare amount of overland flow due to saturation excess. The fast response mechanisms, both in magnitude and relative runoff contribution, were unprecedented and can be considered as extreme.

The magnitudes of the precipitation, flow mechanisms and discharge were extreme for the 2021 flood event. The extreme flows were possible due to fast flow mechanisms, most likely overland flow due to saturation excess, which can be linked to the extreme pre-conditions. The strong relief and impermeable soils also seem to play a role in this extreme response. All catchments are capable of a fast and strong runoff response and although the forcing of the event was extreme, a new flood event in the future can not be excluded and so a higher level of preparedness and protection is required.

This study should be considered as an exploratory first step for action on the flood event. In order to move forward and increase the protection of the region, more detailed research is required. Such research can focus on several areas. Varying the study regions could be helpful for specific purpose. Several examples are an analysis of the entire region for a precipitation analysis, one catchment study for flood measures, or a catchment comparison for detecting hydrological influences. The anthropogenic influences can also provide more insight and opportunities and could be included in these future researches. The use of models was limited in this research but their further development is essential for flood risk management. Uncertainty is another topic that deserves more attention. An important factor in decreasing the uncertainty is to construct robust and extensive measurement infrastructure. Nevertheless, without more international cooperation and data accessibility the additional data will not reach its full potential. International cooperation also provides more value to the previous recommendations. This research shows that the transboundary catchments of the case study can be compared and give more insights to the overall behaviour.