Recent  

In this study, three machine learning methods displayed a good detection capacity of flash droughts. The RF model was recommended to estimate the depletion rate of soil moisture and simulate flash drought by considering the multiple meteorological variable anomalies in the adjacent time to drought onset. The anomalies of precipitation and potential evapotranspiration exhibited a stronger synergistic but asymmetrical effect on flash droughts compared to slowly developing droughts.

In the study we analysed the uncertainties of the meteorological data and model parameterization for evaporation modelling. We have taken a physically based lumped BROOK90 model and applied it in three different frameworks using global, regional and local datasets. Validating the simulations with eddy-covariance data from five stations in Germany, we found that the accuracy model parameterization plays a bigger role than the quality of the meteorological forcing.

Reliable data on global inundated areas remain uncertain. By matching a leading global data product on inundation extents (GIEMS) against predictions from a global hydrodynamic model (CaMa-Flood), we found small but consistent and non-random biases in well-known tropical wetlands (Sudd, Pantanal, Amazon and Congo). These result from known limitations in the data and the models used, which shows us how to improve our ability to make critical predictions of inundation events in the future.

In hydrology the formation of landform patterns is of special interest as changing forcings of the natural systems, such as climate or land use, will change these structures. In our study we developed a thermodynamic framework for surface runoff on hillslopes and highlight the differences of energy conversion patterns on two related spatial and temporal scales. The results indicate that surface runoff on hillslopes approaches a maximum power state.

Regional climate models not only provide projections on temperature and precipitation, but also on snow. Here, we employed statistical post-processing using satellite observations to reduce bias and uncertainty from model projections of future snow-covered area and duration under different greenhouse gas concentration scenarios for the European Alps. Snow cover area/duration decreased overall in the future, three times more strongly with 4–5° global warming as compared to 1.5–2°.

Despite the accuracy of deep learning rainfall-runoff models, we are currently uncertain of what these models have learned. In this study we explore the internals of one deep learning architecture and demonstrate that the model learns about intermediate hydrological stores of soil moisture and snow water, despite never having seen data about these processes during training. Therefore, we find evidence that the deep learning approach learns a physically realistic mapping from inputs to outputs.

We develop a Gaussian Bayesian network (GBN) for seasonal forecasting of lake water quality and algal bloom risk in a nutrient-impacted lake in southern Norway. Bayesian networks are powerful tools for environmental modelling but are almost exclusively discrete. We demonstrate that a continuous GBN is a promising alternative approach. Predictive performance of the GBN was similar or improved compared to a discrete network, and it was substantially less time-consuming and subjective to develop.

In this study, different extensions of the degree-day model were calibrated on snow-cover distribution against freely available satellite snow-cover images. The calibrated models simulated the distribution very well in Baden-Württemberg (Germany) and Switzerland. In addition to reliable identification of snow cover, the melt outputs from the calibrated models were able to improve the flow simulations in different catchments in the study region.

This study explored the quantitative contribution of agricultural intensification and climate change to the sediment load of a small agricultural watershed. Rather than a change in climatic conditions, changes in the land structure notably altered sediment concentrations under high-flow conditions, thereby contributing most to the increase in annual sediment loads. More consideration of land structure improvement is required when combating the transfer of soil from land to water.

This study investigates spatial and temporal patterns in the incidence angle dependence of backscatter from the ASCAT C-band scatterometer and relates those to precipitation, humidity, and radiation data and GRACE equivalent water thickness in ecoregions in the Amazon. The results show that the ASCAT data record offers a unique perspective on vegetation water dynamics exhibiting sensitivity to moisture availability and demand and phenological change at interannual, seasonal, and diurnal scales.

The European Flood Awareness System creates flood forecasts for up to 15 d in the future for the whole of Europe which are made available to local authorities. These forecasts can be erroneous because the weather forecasts include errors or because the hydrological model used does not represent the flow in the rivers correctly. We found that, by using recent observations and a model trained with past observations and forecasts, the real-time forecast can be corrected, thus becoming more useful.

How to combine multi-source precipitation data effectively is one of the hot topics in hydrometeorological research. This study presents a two-step merging strategy based on machine learning for multi-source precipitation merging over China. The results demonstrate that the proposed method effectively distinguishes the occurrence of precipitation events and reduces the error in precipitation estimation. This method is robust and may be successfully applied to other areas even with scarce data.

Precipitation forecasting has potential uncertainty due to data and model uncertainties. Here, an integrated predictive uncertainty modeling framework is proposed by jointly considering data and model uncertainties through an uncertainty propagation theorem. The results indicate an effective predictive uncertainty estimation for precipitation forecasting, indicating the great potential for uncertainty quantification of numerous predictive applications.

Many western Africans encounter serious floods every year. The FANFAR project co-designed a pre-operational flood forecasting system (FEWS) with 50 key western African stakeholders. Participatory multi-criteria decision analysis (MCDA) helped prioritize a FEWS that meets their needs: it should provide accurate, clear, and timely flood risk information and work reliably in tough conditions. As a theoretical contribution, we propose an assessment framework for transdisciplinary hydrology research.

We studied the controls on open water evaporation with a focus on Lake IJssel, the Netherlands, by analysing eddy covariance observations over two summer periods at two locations at the borders of the lake. Wind speed and the vertical vapour pressure gradient can explain most of the variation in observed evaporation, which is in agreement with Dalton's model. We argue that the distinct characteristics of inland waterbodies need to be taken into account when parameterizing their evaporation.

Different models were developed to increase the temporal resolution of precipitation time series to minutes. Their applicability under climate change and their suitability for producing input time series for green infrastructure (e.g. green roofs) modelling were evaluated. The robustness of the model was validated against a range of European climates in eight locations in France and Norway. The future hydrological performances of green roofs were evaluated in order to improve design practice.

Aquifers often act as low-pass filters, dampening high-frequency (intra-annual) and amplifying low-frequency (LFV, multi-annual to multidecadal) variabilities originating from climate variability. By processing groundwater level signals, we show the key role of LFV in the occurrence of groundwater extremes (GWEs). Results highlight how changes in LFV may impact future GWEs as well as the importance of correct representation of LFV in general circulation model outputs for GWE projection.

In this paper, we present an application of the electromagnetic method to image the subsurface below rivers, lakes, or any surface water body. The scanning of the subsurface is carried out by sailing an electromagnetic sensor called FloaTEM. Imaging results show a 3D distribution of different sediment types below the freshwater lakes. In the case of saline water, the system is capable of identifying the probable location of groundwater discharge into seawater.

Climate change affects precipitation phase, which can propagate into changes in streamflow timing and magnitude. This study examines how variations in rainfall and snowmelt affect discharge. We found that annual discharge and stream cessation depended on the magnitude and timing of rainfall and snowmelt and on the snowpack melt-out date. This highlights the importance of precipitation timing and emphasizes the need for spatiotemporally distributed simulations of snowpack and rainfall dynamics.

Reliable prediction of floods depends on the quality of the input data such as precipitation. However, estimation of precipitation from the local measurements is known to be difficult, especially for extremes. Regionalization improves the estimates by increasing the quantity of data available for estimation. Here, we compare three regionalization methods based on their robustness and reliability. We apply the comparison to a dense network of daily stations within and outside Switzerland.