Flooding is a frequent disaster that has a wide-spread footprint globally with significant financial and societal impacts. With availability of Earth observation data from private and public entities at varying spatial, temporal, and spectral resolution as well as data from crowdsourcing, there is no shortage of models. In fact, models and algorithms are abundant and proliferating. However, the question remains where is a global flood model when we need one? Just because models are available does not mean they are usable or accessible and adequate for emergency managers, first responders and other stakeholders who use the model outputs for preparedness, response and resource planning. Often the issue of usability stems from the fact that the models are not always reproducible or replicable. The accuracy and uncertainty associated with the models and how they change based on the scale of analysis and the resolution of input and output datasets are often not communicated properly to stakeholders so they can be part of their decision-making process. The proliferation of machine learning and data driven models that rely on historical data also adds to this problem. This paper discusses several important issues associated with global flood models and provides recommendations that could be used to increase the usability of these models.
Forested watersheds provide many ecosystem services that have become increasingly threatened by wildfire. Stream nitrate (NO 3 -) concentrations often increase following wildfire and can remain elevated for decades. We investigated the drivers of persistent elevated stream NO 3 - in nine watersheds that were burned to varying degrees 16 years prior by the Hayman fire, Colorado, USA. We evaluated the ability of multiple linear regression and spatial stream network modeling approaches to predict observed concentrations of the biologically active solute NO 3 - and the conservative solute sodium (Na +). Specifically, we quantified the degree to which landscape and stream network characteristics predict stream solute concentrations. No landscape variables were strong predictors of stream Na +. Rather, stream Na + variability was largely attributed to flow-connected spatial autocorrelation, indicating that downstream hydrologic transport was the primary driver of spatially distributed Na + concentrations. In contrast, vegetation cover, measured as mean normalized differenced moisture index (NDMI), was the strongest predictor of spatially distributed stream NO 3 - concentrations. Furthermore, stream NO 3 - concentrations had weak flow-connected spatial autocorrelation and high spatial variability. This pattern is likely the result of spatially heterogeneous wildfire behavior that leaves intact forest patches interspersed with high burn severity patches that are dominated by shrubs and grasses. Post-fire vegetation also interacts with watershed structure to influence stream NO 3 - patterns. For example, severely burned convergent hillslopes in headwaters positions were associated with the highest stream NO 3 ‑ concentrations due to the high proportional influence of hillslope water in these locations. Our findings suggest that reforestation is critical for the recovery of stream NO 3 - concentrations to pre-fire levels and targeted planting in severely burned convergent hillslopes in headwater positions will likely have a large impact on stream NO 3 - concentrations.
Ecohydrological investigations commonly use the stable isotopes of water (hydrogen and oxygen) as a conservative ecosystem tracer. This approach requires accessing and analyzing water constrained within plant and soil matrices. Generally, there are six steps that researchers must pass through to retrieve hydrogen and oxygen isotope values from these plant and soil matrices: (i) sampling, (ii) sample storage and transport, (iii) extraction, (iv) pre-analysis processing, (v) isotopic analysis, and (vi) post-processing and correction. At each of these steps cumulative errors can be introduced which sum to non-trivial magnitudes. These errors can impact subsequent interpretations about water cycling through the soil-plant-atmosphere continuum. But these steps in the research ‘process chain’ are just the tip of the iceberg when it comes to uncertainly in published findings. At each of these discreet steps, there are multiple possible options to select from resulting in, as we will show, tens of thousands of possible combinations used by researchers to go from plant and soil samples to isotopic data. In a newly emerging science, so many options can create interpretive confusion and major issues with data comparability. This points to the need for the development of shared standardized approaches. Here we critically examine the state of the process chain, reflecting on the issues associated with each step, and end with suggestions to move our community towards standardization. We hope that critically assessing this common approach will help us see the current problem in its entirety and facilitate community action toward agreed upon standardized approaches.
Using annual water balance analyses may mask intra-annual variability in runoff generation, which could limit our understanding of the similarities and differences between water- and energy-limited catchments. This may be especially limiting in comparisons between catchments close to the threshold between water- and energy-limitation. For this study, we examined runoff generation as a function of catchment storage in four watersheds, with focus on two that exist close to these thresholds to identify how year-to-year variability in storage resulted in intra-annual variations of runoff generation efficiency. Specifically, we focused on one energy-limited catchment in the humid subtropics and one water-limited in a Mediterranean climate. We used measured and calculated daily water balance components to calculate variations in the relative magnitude of daily storage. We isolated precipitation events to draw connections between storage and runoff generation at intra-annual scales and compared our findings to the same metrics in two intensely energy-limited landscapes. We observed distinct stages in daily storage across water years in watersheds at the threshold, where systems experienced wet-up, plateau, and dry-down stages. During the wet-up, precipitation was partitioned to storage, and runoff ratios ( RR) were low. In the plateau, storage was filled, precipitation was partitioned to runoff, causing high RRs. During the dry-down, storage decreased as precipitation was partitioned to evapotranspiration and runoff, causing low RRs. The critical role of evapotranspiration during the growing season resulted in relatively higher RRs during the wet-up than during the dry-down for a given storage value. Thus the same storage amount was partitioned to evapotranspiration or runoff differently throughout the year, depending on the storage stage. Despite their different positions on opposite sides of the threshold, the similarity between the two focus catchments suggests a potential characteristic behavior of systems at the threshold common to both humid and semi-arid landscapes.
Insight into the rainfall-soil moisture (SM) response to land cover is critical for soil hydrological process modeling and management. In this study, five typical land-cover types (forest, shrub, grass, crop, and bare land) and four rainfall patterns (heavy, intermediate, light, and continuous rains) were selected to assess the effects of SM response characteristics on the Loess Plateau of China. We monitored SM at five depths on each land-cover type at 1-h intervals over the growing season of 2019. The results showed that rainfall patterns and land-cover typestogether determined the SM response process and infiltration efficiency. A minimum accumulated rainfall amount of 5 mm was the threshold to trigger a 10-cm SM response. Rain events with higher intensity and smaller sum triggered a quick surface SM response, while larger amounts could percolate deeper and faster. Land-cover change significantly altered the rainfall-SM response dynamics and rainwater utilization efficiency after 20 years of ecological construction. Revegetation sites (mean values of forest, shrub, and grass) increased the soil wetting depth by 14.7%, shortened the SM response time by 27.3%, and accelerated the SM wetting front velocity by 67.2%, which promoted a 35.2% rainfall transformation rate (RTR) across the 1-m profile over all rainfall events (R 1-13). Moreover, planted forest showed the highest RTR of R 1-13 and the maximal increase in soil water storage, which did not aggravate the soil water deficit across the 1-m profile over the growing season. Therefore, we present evidence that planted forests, instead of shrubs, may be beneficial for water conservation if precipitation is greater than 550 mm. The findings of this study prove the role of revegetation on rainwater infiltration capacity and efficiency and can help improve the management of afforestation in arid and semiarid regions.
The projected climate change for Norway through the 21st century predicts that the temperature will increase significantly. Events with heavy rainfall will be more intense and occur more frequently. Rain floods will increase in magnitude and also occur more frequently. Extreme flooding and heavy rain will significantly impact the sediment dynamics in rivers. In the mountain areas, floods are often associated with erosion, transport and deposition of coarse sediment along the streams. These processes are related to bed load transport and pose a hazard in addition to the elevated water discharge and have to be included in management plans for river basins. This paper studies the bed load delivery from sources that contribute the most to the sediment budget in the Gudbrandsdalslågen river basin during the large magnitude floods in 2011 and 2013. More than 100 debris slides and debris flow were triggered in the tributary river Veikleåi by the heavy rain and snowmelt during these floods. The volume of the contribution from debris flows and erosion and deposition of the river bed was determined by subtracting digital elevation models acquired during repeated airborne LIDAR surveys. In the river Dørja the supply of sediment from a number of debris flows caused extensive aggradation and channel changes. In their new position, lateral erosion by these channels triggered slides on the adjacent slopes. The contributing volumes of debris flows, lateral erosion, and river-bed erosion and deposition were determined from the LIDAR surveys. Relations obtained from studies of sediment transport in modern glacier rivers were used to obtain estimates of the ratio of bed load versus suspended load derived from the Pleistocene moraine deposits. Several monitoring stations using conventional methods for measuring bed load and suspended load recorded very large volumes of sediment delivery during both of the extreme floods. The results of the study may be used to identify necessary locations for check-dams and erosion protection works in order to adapt to climate change.
In many parts of South Africa, soil erosion rates are high, and likely to be exacerbated by the longer droughts and more intense rainfall that are predicted in long-term regional climate change scenarios. Suspended sediment loads (SSL) and yields (SSY) are accepted means of expressing and comparing sediment transport and soil erosion rates. Land care and water security initiatives in South Africa require these data to provide benchmarking, and trajectories of change. International researchers began in the 1970s to investigate SSL estimation approaches. These investigations typically used near-continuous turbidity data from installed probes as a surrogate for sampled SS, and auto-samplers to monitor SS concentration and develop sediment rating curves. Biophysical and socio-economic conditions in South Africa differ markedly from the northern hemisphere environments where foundational studies were conducted. SSL estimations in South Africa are associated with extreme hydrological regimes, remote study areas and lack the resources required to collect and analyse representative SS data. There is a dearth of measured SS data, and of observed SSL and SSY for South African catchments. Using measured SS data from the Tsitsa River catchment (Eastern Cape, South Africa) we found that a discharge-weighted interpolation estimator was more appropriate than regression estimators, and that SSY responses to biophysical factors were in some ways more similar to northern hemisphere norms than expected. Lack of technical, infrastructural, human and financial resources were our main constraints to monitoring and estimating SSY. Our findings highlight the challenges of, and provide some guidance for, estimating directly measured SSL in the southern Africa region and inform future research in resource scarce areas.
River estuaries are characterized by mixing processes between marine water masses and freshwater inflows depleted in heavier stable isotopes. Therefore, they often show a linear correlation between salinity and water isotopes (δ 18O and δ 2H values). In this study we evaluated spatial and seasonal isotope dynamics along three estuarine lagoon transects, located at the northern German Baltic Sea coast. The data show strong seasonality of isotope values, even at locations located furthest from the river mouths. They further reveal a positive and linear salinity-isotope correlation in spring, but hyperbolic and partially reverse correlations in summers. We conclude, that additional physical processes such as evaporation from the shallow lagoons, partially overprint the two-phase mixing correlation during summers. Understanding those water isotope and salinity dynamics are crucial in context of ecological studies. For example, when interpreting oxygen and hydrogen isotope data in aquatic organisms, that depend on ambient estuarine waters.
Smart drainage management to limit summer drought damage in Nordic agriculture under the circular economy conceptSyed Md Touhidul Mustafa 1, *, Kedar Ghag1, Anandharuban Panchanathan2, Bishal Dahal 1, Amirhossein Ahrari1, Toni Liedes 3, Hannu Marttila1, Tamara Avellán1, Mourad Oussalah2, Björn Klöve 1, & Ali Torabi Haghighi11Water, Energy and Environmental Engineering Research Unit, University of Oulu, P.O. Box 4300, FIN90014, Oulu, Finland2Center for Machine Vision and Signal Analysis, University of Oulu, Finland3Intelligent Machines and Systems, University of Oulu, PO Box 4200, 90014 Oulu, Finland
It has been almost 50 years since Mike Bonell’s foundational work in the humid tropics, kickstarting the field of tropical hydrology. In order to expand on this work and build a more generalized hydrological understanding of steep rainforest catchments, we studied the seasonal and inter-annual evolution of hydrological response from two catchments with similar characteristics to those studied by Bonell. Both hydrometric and water stable isotope data were collected at relatively high frequencies during one wet season (Thompson Creek) and a three-year period (Atika Creek). The longer dataset spans a wide range of environmental conditions experienced in the humid tropics, including events that cover the wetting-up transitional period of the wet season, ENSO events and tropical cyclones. Both catchments displayed fast streamflow response to rainfall with the shallow upper soil profile responding quickly to rainfall at Atika Creek. New findings from this study include the importance of pre-event water (>50%) for overall event flows, especially when the catchment was wet. Rainfall, surface runoff and groundwater isotope compositions varied between rainfall events with the most complex mixing plots observed for multi-peak events that occurred at the start of the wet season and after a dry period within the wet season. Inter-annual variability in catchment hydrology reflected changing ENSO conditions and the 2020-21 La Ninã wet season was characterized by several tropical cyclone events which generated the most 18O-depleted rainfall and streamflow isotope values. Our findings highlight the requirement for high frequency multi-source sampling to accurately interpret catchment behavior. We propose a conceptual model to describe the seasonal evolution of streamflow response in steep rainforest catchments.
Modelling the response of hydrological processes to the changing climate requires the use of a chain of numerical models, each of which contributes some degree of uncertainty to the final outputs. As a result, hydrological projections, despite the progressive increase in the accuracy of the models along the chain, can still display high levels of uncertainty, especially at small temporal and spatial scales. The randomness intrinsic to climate phenomena, known as internal climate variability, is also a component contributing to the uncertainty of the hydrological projections. Unlike the uncertainties emerging from the climate and hydrological models, the internal climate variability is irreducible. In this work, we quantify and partition the uncertainty of hydrological processes in two mountainous catchments in Switzerland, emerging from climate models and internal variability, across a broad range of scales. To that end, we used high-resolution ensembles of climate and hydrological data, produced by a two-dimensional weather generator (AWE-GEN-2d) and a distributed hydrological model (Topkapi-ETH). We quantified the uncertainty in hydrological projections towards the end of the century through the estimation of the values of signal-to-noise ratios (STNR). We found small STNR values (<-1) in the projection of annual streamflow for most sub-catchments in both study sites that are dominated by the large natural variability of precipitation (explains ~70% of total uncertainty). Furthermore, we investigated in detail specific hydrological components that are critical in the model chain. For example, snowmelt or liquid precipitation exhibits robust change signals, which translates into high STNR values for streamflow during warm seasons and at higher elevations, together with a larger contribution of climate model uncertainty, suggesting that an improvement of the involved models has the potential of significantly narrowing the uncertainty. In contrast, extreme flows show low STNR values due to large internal climate variability across all elevations, which limits the possibility of narrowing their estimation uncertainty in a warming climate.
Process understanding of the interaction between streamflow, groundwater and water usages under drought are hampered by a limited number of gauging stations, especially in tributaries. Recent technological advances facilitate the application of non-commercial measurement devices for monitoring environmental systems. The Dreisam River in the South-West of Germany was affected by several hydrological drought events from 2015 to 2020, when parts of the main stream and tributaries fell dry. A flexible longitudinal water quality and quantity monitoring network was set up in 2018. Among other measurements it employs an image based method with QR codes as fiducial marker. In order to assess under which conditions the QR-code based water level loggers (WLL) deliver data according to scientific standards, we present a comparison to conventional capacitive based WLL. The results from 20 monitoring stations reveal that the riverbed was dry for > 50 \% at several locations and even for > 70 \% at most severely affected locations during July and August 2020, with the north western parts of the catchment being especially concerned. Thus, the highly variable longitudinal drying patterns of the stream reaches could be monitored. The image-based method was found to be a valuable asset for identification of confounding factors and validation of zero level occurrences. Nevertheless, a simple image processing approach (based on an automatic thresholding algorithm) did not compensate for errors due to natural conditions and technical setup. Our findings highlight that the complexity of measurement environments is a major challenge when working with image-based methods.
We examined changes in catchment-scale annual and seasonal evapotranspiration after 50% strip thinning, using runoff data from headwater catchments. The short-term water balance (STWB) method between periods from 8 to 100 days was applied to the treated (KT) and control (KC) catchments. The estimated evapotranspiration during the pre- and post-thinning periods were 840 and 910 and 780 and 860 mm/year in the catchments KT and KC, respectively. According to a paired catchment analysis of estimated evapotranspiration, monthly evapotranspiration increased from 3 to 20 mm from June to December, while it decreased from 7 to 31 mm from January to May. The estimated annual and monthly evapotranspiration was compatible with the values monitored in the plot-scale interception, canopy transpiration, and ground surface evapotranspiration. Our findings showed that the decreases in evapotranspiration due to 50% thinning were similar or different in different methods of measurement when compared with thinning in the other catchments around the world. The STWB model can evaluate the effects of timber harvesting on changes in evapotranspiration (ET), including the reproduction of seasonal patterns of ET.
Concentration-discharge (C-Q) relationships can provide insight into how catchments store and transport solutes, but analysis is often limited to long-term behavior assessed from infrequent grab samples. Increasing availability of high-frequency sensor data has shown that C-Q relationships can vary substantially across temporal scales, and in response to different hydrologic drivers. Here, we present four years of dissolved organic carbon (DOC) and nitrate-nitrogen (NO3-N) sensor data from a snowmelt- dominated catchment in the Rocky Mountains of Colorado. We assessed both the direction (enrichment vs. dilution) and hysteresis in C-Q relationships across a range of time scales, from interannual to sub-daily. Both solutes exhibited a seasonal flushing response, with concentrations initially increasing as solute stores are mobilized by the melt pulse, but then declining as these stores are depleted. The high-frequency data revealed that the seasonal melt pulse was composed of numerous individual daily melt pulses. The solute response to daily melt pulses was relatively chemostatic, suggesting mobilization and depletion to be progressive rather than episodic processes. In contrast, rainfall-induced pulses produced short-lived but substantial enrichment responses, suggesting they may activate alternative solute sources or transport pathways. The results clearly demonstrate that solute responses to individual events may differ significantly from the longer-term behavior these events combine to generate, something which only becomes apparent when high-frequency data are used.
Accurate simulation of plant water use across agricultural ecosystems is essential for various applications, including precision agriculture, quantifying groundwater recharge, and optimizing irrigation rates. Previous approaches to integrating plant water use data into hydrologic models have relied on evapotranspiration (ET) observations. Recently, the flux variance similarity approach has been developed to partition ET to transpiration (T) and evaporation, providing an opportunity to use T data to parameterize models. To explore the value of T/ET data in improving hydrologic model performance, we examined multiple approaches to incorporate these observations for vegetation parameterization. We used ET observations from 5 eddy covariance towers located in the San Joaquin Valley, California, to parameterize orchard crops in an integrated land surface – groundwater model. We find that a simple approach of selecting the best parameter sets based on ET and T performance metrics works best at these study sites. Selecting parameters based on performance relative to observed ET creates an uncertainty of 27% relative to the observed value. When parameters are selected using both T and ET data, this uncertainty drops to 24%. Similarly, the uncertainty in potential groundwater recharge drops from 63% to 58% when parameters are selected with ET or T and ET data, respectively. Additionally, using crop type parameters results in similar levels of simulated ET as using site-specific parameters. Different irrigation schemes create high amounts of uncertainty and highlight the need for accurate estimates of irrigation when performing water budget studies.
Soil water repellency (SWR) increases surface runoff and preferential flows. Thus, quantitative evaluation of SWR distribution is necessary to understand water movements. Because the variability of SWR distribution makes it difficult to measure directly, we developed a method for estimating an SWR distribution index, defined as the areal fraction of surface soil showing SWR (SWRarea). The theoretical basis of the method is as follows: (1) SWRarea is equivalent to the probability that a position on the soil surface is drier than the critical water content (CWC); SWR is present (droplets absorbed in >10 s) when the soil surface is drier than the CWC and absent when it is wetter. (2) CWC and soil moisture content (θ) are normally distributed independent variables. (3) Thus, based on probability theory, the cumulative normal distribution of θ – CWC (f(x)) can be obtained from the distributions of CWC and θ, and f(0), the cumulative probability that θ – CWC < 0, gives the SWRarea. To investigate whether the method gives reasonable results, we repeatedly measured θ at 0–5 cm depth and determined the water repellency of the soil surface at multiple points in fixed plots with different soils and topography in a humid-temperate forest. We then calculated the CWC from the observed θ–SWR relationship at each point. We tested the normality of the CWC and θ distributions and the correlation between CWC and θ. Then, we determined f(x) from the CWC and θ distributions and estimated the SWRarea on each measurement day. Although CWC and θ were both normally distributed, in many cases they were correlated. Nevertheless, the CWC–θ dependency had little effect on the estimation error, and f(x) explained 69% of the SWRarea variability. Our findings show that a stochastic approach is useful for estimating SWRarea.
Lateral saturated soil hydraulic conductivity, Ks,l, is the soil property governing subsurface water transfer in hillslopes, and the key parameter in many numerical models simulating hydrological processes both at the hillslope and catchment scales. Likewise, the hydrological connectivity of lateral flow paths plays a significant role in determining the intensity of the subsurface flow at various spatial scales. The objective of the study is to investigate the relationship between Ks,l and hydraulic connectivity at the hillslope spatial scale. Ks,l was determined by the subsurface flow rates intercepted by drains, and by water table depths observed in a well network. Hydraulic connectivity of the lateral flow paths was evaluated by the synchronicity among piezometric peaks, and between the latter and the peaks of drained flow. Soil moisture and precipitation data were used to investigate the influence of the transient hydrological soil condition on connectivity and Ks,l. It was found that the higher was the synchronicity of the water table response between wells, the lower was the time lag between the peaks of water levels and those of the drained subsurface flow. Moreover, the most synchronic water table rises determined the highest drainage rates. The relationships between Ks,l and water table depths were highly non-linear, with a sharp increase of the values for water table levels close to the soil surface. Estimated Ks,l values for the full saturated soil were in the order of thousands of mm h-1, suggesting the activation of macropores in the root zone. The Ks,l values determined at the peak of the drainage events were correlated with the indicators of synchronicity. The sum of the antecedent soil moisture and of the precipitation was correlated with the indicators of connectivity and with Ks,l. We suggest that, for simulating realistic processes at the hillslope scale, the hydraulic connectivity could be implicitly considered in hydrological modelling through an evaluation of Ks,l at the same spatial scale.