Hydroclimatic extremes such as intense rainfall, floods, droughts, heatwaves, and wind or storms have devastating effects each year. One of the key challenges for society is understanding how these extremes are evolving and likely to unfold beyond their historical distributions under the influence of multiple drivers such as changes in climate, land cover, and other human factors. Methods for analysing hydroclimatic extremes have advanced considerably in recent decades. Here we provide a review of the drivers, metrics, and methods for the detection, attribution, management, and projection of nonstationary hydroclimatic extremes. We discuss issues and uncertainty associated with these approaches (e.g. arising from insufficient record length, spurious nonstationarities, or incomplete representation of nonstationary sources in modelling frameworks), examine empirical and simulation-based frameworks for analysis of nonstationary extremes, and identify gaps for future research.
21 October 2021 in Hydrology and Earth System Sciences
Long short-term memory (LSTM) models are recurrent neural networks from the field of deep learning (DL) which have shown promise for time series modelling, especially in conditions when data are abundant. Previous studies have demonstrated the applicability of LSTM-based models for rainfall–runoff modelling; however, LSTMs have not been tested on catchments in Great Britain (GB). Moreover, opportunities exist to use spatial and seasonal patterns in model performances to improve our understanding of hydrological processes and to examine the advantages and disadvantages of LSTM-based models for hydrological simulation. By training two LSTM architectures across a large sample of 669 catchments in GB, we demonstrate that the LSTM and the Entity Aware LSTM (EA LSTM) models simulate discharge with median Nash–Sutcliffe efficiency (NSE) scores of 0.88 and 0.86 respectively. We find that the LSTM-based models outperform a suite of benchmark conceptual models, suggesting an opportunity to use additional data to refine conceptual models. In summary, the LSTM-based models show the largest performance improvements in the north-east of Scotland and in south-east of England. The south-east of England remained difficult to model, however, in part due to the inability of the LSTMs configured in this study to learn groundwater processes, human abstractions and complex percolation properties from the hydro-meteorological variables typically employed for hydrological modelling.
10 January 2022 in Communications Earth & Environment
Ambitious afforestation proposals in the last decade target potential flood mitigation and carbon storage benefits but without a systematic, large-scale (>1000 km2) quantitative evaluation of their impacts on streamflow. Here, we assess the impact of afforestation on streamflow across twelve diverse catchments (c.500-10,000 km2) using a high-resolution land-surface model with a large ensemble of afforestation scenarios. Afforestation consistently decreases median and low streamflow. Median modelled flow is reduced by 2.8% ± 1.0 (1 s.d.), or 10 mm yr−1 ± 2.1 (1 s.d.), for a ten-percentage point increase in catchment broadleaf woodland. We find no nationally-consistent reduction of extreme floods. In larger catchments, planting extent is a stronger control on streamflow than location. Our results suggest that despite its potential environmental and societal benefits, widespread afforestation may inadvertently reduce water availability, particularly in drier areas, whilst only providing a modest reduction in extreme flood flows.
Widespread afforestation has been proposed internationally to reduce atmospheric carbon dioxide; however, the specific hydrological consequences and benefits of such large-scale afforestation (e.g. natural flood management) are poorly understood. We use a high-resolution land surface model, the Joint UK Land Environment Simulator (JULES), with realistic potential afforestation scenarios to quantify possible hydrological change across Great Britain in both present and projected climate. We assess whether proposed afforestation produces significantly different regional responses across regions; whether hydrological fluxes, stores and events are significantly altered by afforestation relative to climate; and how future hydrological processes may be altered up to 2050. Additionally, this enables determination of the relative sensitivity of land surface process representation in JULES compared to climate changes. For these three aims we run simulations using (i) past climate with proposed land cover changes and known floods and drought events; (ii) past climate with independent changes in precipitation, temperature, and CO2; and (iii) a potential future climate (2020–2050). We find the proposed scale of afforestation is unlikely to significantly alter regional hydrology; however, it can noticeably decrease low flows whilst not reducing high flows. The afforestation levels minimally impact hydrological processes compared to changes in precipitation, temperature, and CO2. Warming average temperatures (+3 °C) decreases streamflow, while rising precipitation (130 %) and CO2 (600 ppm) increase streamflow. Changes in high flow are generated because of evaporative parameterizations, whereas low flows are controlled by runoff model parameterizations. In this study, land surface parameters within a land surface model do not substantially alter hydrological processes when compared to climate.
Afforestation is of international interest for its positive benefits on carbon storage, ecology, and society, but its impacts on terrestrial and atmospheric processes are still poorly understood. This study presents the first use of a coupled land surface and convection permitting atmospheric model (CPM) to quantify hydrometeorological effects of afforestation across the United Kingdom, focusing on atmospheric processes often missing in hydrological models. Generating a scenario of 93 000 km2 (40%) additional woodland across the UK, the periods of 2042–2052 and 2062–2072 are analysed. Simulated afforestation alters seasonal and regional UK hydrometeorology. Countrywide runoff increases in all seasons (between 5.4–11 mm and 4.3–8.6% per season) due to elevated subsurface flows from greater soil moisture. Evaporation decreases in summer (−20.6 mm, −10%) but increases in winter (8.1 mm, 15%) whereas rainfall increases throughout all seasons (between 2.2–6.86 mm and 0.9%–2.2% per season). Greater winter rainfall is detected along Great Britain’s west coastline as increased surface roughness produces prolonged and heavier rainfall. In the summer, lower albedo increases potential evapotranspiration and reduces near surface specific humidity: water is locked in deeper soil layers as transpiration diminishes and the topsoil dries out. However, the magnitude of hydrometeorological change due to altered land cover is smaller than the uncertainty in local climate change projections. This work sets a precedent in illustrating the impacts of afforestation on hydrology using a high-resolution CPM and highlights the importance of coupled hydrometeorological processes when investigating land cover impacts on hydrological processes.
Hydroclimatic extremes such as intense rainfall, floods, droughts, heatwaves, and wind/storms have devastating effects each year. One of the key challenges for society is understanding how these extremes are evolving and likely to unfold beyond their historical distributions under the influence of multiple drivers such as changes in climate, land cover, and other human factors. Methods for analysing hydroclimatic extremes have advanced considerably in recent decades. Here we provide a review of the drivers, metrics and methods for the detection, attribution, prediction and projection of nonstationary hydroclimatic extremes. We discuss issues and uncertainty associated with these approaches (e.g arising from insufficient record length, spurious nonstationarities, or incomplete representation of nonstationary sources in modelling frameworks), examine empirical and simulation-based frameworks for analysis of nonstationary extremes, and identify gaps for future research.