Choosing between post-processing precipitation forecasts or chaining several uncertainty quantification tools in hydrological forecasting systems

This study aims to decipher the interactions of a precipitation post-processor and several other tools for uncertainty quantification implemented in a hydrometeorological forecasting chain. We make use of four hydrometeorological forecasting systems that differ by how uncertainties are estimated and propagated. They consider the following sources of uncertainty: system A, forcing, system B, forcing and initial conditions, system C, forcing and model structure, and system D, forcing, initial conditions, and model structure. For each system's configuration, we investigate the reliability and accuracy of post-processed precipitation forecasts in order to evaluate their ability to improve streamflow forecasts for up to 7 d of forecast horizon. The evaluation is carried out across 30 catchments in the province of Quebec (Canada) and over the 2011–2016 period. Results are compared using a multicriteria approach, and the analysis is performed as a function of lead time and catchment size. The results indicate that the precipitation post-processor resulted in large improvements in the quality of forecasts with regard to the raw precipitation forecasts. This was especially the case when evaluating relative bias and reliability. However, its effectiveness in terms of improving the quality of hydrological forecasts varied according to the configuration of the forecasting system, the forecast attribute, the forecast lead time, and the catchment size. The combination of the precipitation post-processor and the quantification of uncertainty from initial conditions showed the best results. When all sources of uncertainty were quantified, the contribution of the precipitation post-processor to provide better streamflow forecasts was not remarkable, and in some cases, it even deteriorated the overall performance of the hydrometeorological forecasting system. Our study provides an in-depth investigation of how improvements brought by a precipitation post-processor to the quality of the inputs to a hydrological forecasting model can be cancelled along the forecasting chain, depending on how the hydrometeorological forecasting system is configured and on how the other sources of hydrological forecasting uncertainty (initial conditions and model structure) are considered and accounted for. This has implications for the choices users might make when designing new or enhancing existing hydrometeorological ensemble forecasting systems.

Modelling STEMI service delivery: a proof of concept study

BackgroundAccess to individual percutaneous coronary intervention (PCI) centres has traditionally been determined by historical referral patterns along arbitrarily defined geographic boundaries. We set out to produce predictive models of ST-elevation myocardial infarction (STEMI) demand and time-efficient access to PCI centres.MethodsTravel times from random addresses to PCI centres in Melbourne, Australia, were estimated using Google map application programming interface (API). Departures at 08:15 and 17:15 were compared with 23:00 to determine the effect of peak hour traffic congestion. Real-world ambulance travel times were compared with estimated travel times using Google map developer software. STEMI incidence per postcode was estimated by merging STEMI incidence per age group data with age group per postcode census data. PCI centre network configuration changes were assessed for their effect on hospital STEMI loading, catchment size, travel times and the number of STEMI cases within 30 min of a PCI centre.ResultsNearly 10% of STEMI cases travelled more than 30 min to a PCI centre, increasing to 20% by modelling the removal of large outer metropolitan PCI centres (p<0.05). A model of 7 PCI centres compared favourably to the current existing network of 11 PCI centres (p=0.18 (afternoon), p=0.5 (morning and night)). The intraclass correlation between estimated travel times and ambulance travel times was 0.82, p<0.001.ConclusionThis paper provides a framework to integrate prehospital environmental variables, existing or altered healthcare resources and health statistics to objectively model STEMI demand and consequent access to PCI. Our methodology can be modified to incorporate other inputs to compute optimum healthcare efficiencies.

New Workflow of Sediment Mass Balancing, from Local Datasets, for Predicting Basin Scale Trends

Objectives/Scope Sediment volumetric budget estimates are very important input parameters for process-based depositional modelling (forward stratigraphic modelling). This paper presents a new integrated approach for analyze sediment volumetric budgets in sedimentary basins that is based on the reconstruction of regional grain size trends. In subsurface studies of sediment routing systems, noticeable uncertainties in estimated total sediment volumes occur when available datasets are limited to local areas that do not cover the entire sediment routing system. These uncertainties also affect models of catchment areas, structural uplift, and denudation rates as well as net:gross predictions. Methods, Procedures, Process The new integrated approach focuses on reconstructing sediment budgets for entire sediment fairways from limited local datasets. It uses a combination of sediment mass balancing and local grain size distributions to predict basin-wide grain size distributions. The comparison of local grain size to fairway-scale grain size trends is key in correcting sediment volumetrics for significantly reduced uncertainties in catchment reconstruction and net:gross ratios predictions at the scale of sediment fairways, sub-basins, prospects and exploration/production fields. Results, Observations, Conclusions The new approach has been applied successfully to two subsurface continental to marine delta systems. They cover periods of approximately 7 My in total and include four limited local areas of interest (AOI). These local AOIs measure 200×200 km, while the entire sub-basin measures 500×800 km. The new approach indicates that only up to 40% of the total sediment volume of each fairway could be captured by previous methodologies with limited local areas of interest. A maximum of 70% of the entire sink sediment volume could be incorporated in local areas of interest. The new approach presented in this paper significantly lowers the uncertainties in sediment volume estimates, depositional rates and lithology distribution input parameters in forward stratigraphic modelling. For the two case studies, previous sediment flux models indicated rates of 10,000 km/Myr. The new integrated approach indicates that sediment flux actually reached 30,000 km/Myr with major implications for sediment distribution, net:gross prediction and catchment size and denudation rates estimates. Novel/Additive Information The new integrated approach reduces uncertainties in catchment size and tectonic exhumation rate estimates for clastic depositional systems. It provides lower uncertainty parameters (sediment volume, source locations, sediment fractions, diffusion coefficients) for forward stratigraphic modelling, e.g., for reservoir quality prediction in hydrocarbon exploration. In fundamental research, provenance analyses can be better constrained by improved catchment size prediction and sediment grain size distribution models for sink areas

Aluminum in Bottom Sediments of the Lower Silesian Rivers Supplying Dam Reservoirs vs. Selected Chemical Parameters

The study was carried out on sediments collected from three rivers: Nysa Szalona, Strzegomka and Bystrzyca flowing in southwestern Poland. The content of Al in sediments and in bottom water was determined in relation to chemical conditions. The study was carried out in a four-year cycle, during spring and autumn. The aim of the study was to determine the level and accumulation of aluminum in sediments of rivers supplying dam reservoirs storing water for consumption. The sediments studied were mineral in nature, with neutral pH and moderate sulfate content. The level of Al and heavy metals in the sediments was the highest in the Nysa Szalona River and the lowest in the Strzegomka River, which was also evident in the concentration factor (CF). In terms of season, higher Al contents were recorded in sediments in autumn than in spring, which was also reflected in the concentration factor (CF). Along the course of the river, a gradual decrease in Al levels was observed in successive tributaries in the Nysa Szalona and Strzegomka Rivers, while there was no apparent regularity for the Bystrzyca. Against this background, a comparison of extreme sites below the springs and at the reservoir outlet shows that values were higher in the Nysa Szalona below the springs, and lower in the Strzegomka and Bystrzyca below the reservoir outlet. The general picture of Al and heavy metal loading of the studied sediments shows the lowest loading for the Strzegomka, only the enrichment factor (EF) was the lowest for the Nysa Szalona: metal pollution index (MPI)—S < B < NS, contamination factor (Cf)—S < B < NS, degree of contamination (DC)—S < NS < B, EF—NS < B < S, geoaccumulation index (Igeo)—S < B < NS, CF—S < NS < B. There was no effect of catchment size and river length on Al levels in sediments.

Contemporary Sediment Delivery Ratios for Small Catchments Subject to Shallow Rainfall Triggered Earthflows in the Waipaoa Catchment, North Island, New Zealand

<p>The Waipaoa catchment is generally considered to have high hill slope channel coupling due to the large volumes of sediment output at the river mouth. Yet the percentage of sediment that is transported within the fluvial system is low when considered in terms of the total volume of sediment mobilised during episodic failure events. Clearly, there is a discrepancy between generation of sediment and its delivery to the fluvial network. Previous research has suggested there is a strong decrease in catchment connectivity as catchment size increases. However, little research has been undertaken to understand the changes in hillslope-channel coupling over time. This study focuses on the connectivity of shallow rainfall triggered earthflows located in small catchments located within three different land systems in the Waipaoa Catchment. A multiple regression model was developed to predict the sediment delivery ratio for individual earthflows based on an empirical dataset of earthflows which occurred during a storm event in 2002. The results from this modelling were applied to five larger sub-catchments where sequential aerial photograph analysis (1940s to 2004) was used to determine connectivity. From this, spatial and temporal patterns in the catchment sediment delivery ratios were identified. The expected decrease in sediment delivery ratios was observed as catchment size increased. However, the temporal pattern to sediment delivery is not so clear. It appears that catchment evolution, referring specially to the Terrain Event Resistance Model developed by Crozier and Preston (1999), does not have a significant influence on sediment delivery ratios within the six decades examined in this thesis. Furthermore, while earthflows are considered the ultimate source of sediment during storm events, they are not always the mechanism by which this sediment enters the fluvial network. It is also vital to consider rates of gullying, sheet erosion and riparian erosion.</p>

Contemporary Sediment Delivery Ratios for Small Catchments Subject to Shallow Rainfall Triggered Earthflows in the Waipaoa Catchment, North Island, New Zealand

<p>The Waipaoa catchment is generally considered to have high hill slope channel coupling due to the large volumes of sediment output at the river mouth. Yet the percentage of sediment that is transported within the fluvial system is low when considered in terms of the total volume of sediment mobilised during episodic failure events. Clearly, there is a discrepancy between generation of sediment and its delivery to the fluvial network. Previous research has suggested there is a strong decrease in catchment connectivity as catchment size increases. However, little research has been undertaken to understand the changes in hillslope-channel coupling over time. This study focuses on the connectivity of shallow rainfall triggered earthflows located in small catchments located within three different land systems in the Waipaoa Catchment. A multiple regression model was developed to predict the sediment delivery ratio for individual earthflows based on an empirical dataset of earthflows which occurred during a storm event in 2002. The results from this modelling were applied to five larger sub-catchments where sequential aerial photograph analysis (1940s to 2004) was used to determine connectivity. From this, spatial and temporal patterns in the catchment sediment delivery ratios were identified. The expected decrease in sediment delivery ratios was observed as catchment size increased. However, the temporal pattern to sediment delivery is not so clear. It appears that catchment evolution, referring specially to the Terrain Event Resistance Model developed by Crozier and Preston (1999), does not have a significant influence on sediment delivery ratios within the six decades examined in this thesis. Furthermore, while earthflows are considered the ultimate source of sediment during storm events, they are not always the mechanism by which this sediment enters the fluvial network. It is also vital to consider rates of gullying, sheet erosion and riparian erosion.</p>

Choosing between post-processing precipitation forecasts or chaining several uncertainty quantification tools in hydrological forecasting systems

This study aims to decipher the interactions of a precipitation post-processor and several other tools for uncertainty quantification implemented in a hydrometeorological forecasting chain. We make use of four hydrometeorological forecasting systems that differ by how uncertainties are estimated and propagated. They consider the following sources of uncertainty: A) forcing, B) forcing and initial conditions, C) forcing and model structure, and D) forcing, initial conditions, and model structure. For each system's configuration, we investigate the reliability and accuracy of post-processed precipitation forecasts in order to evaluate their ability to improve streamflow forecasts for up to seven days of forecast horizon. The evaluation is carried out across 30 catchments in the Province of Quebec (Canada) and over the 2011–2016 period. Results are compared using a multicriteria approach, and the analysis is performed as a function of lead time and catchment size. The results indicate that the precipitation post-processor resulted in large improvements in the quality of forecasts with regard to the raw precipitation forecasts. This was especially the case when evaluating relative bias and reliability. However, its effectiveness in terms of improving the quality of hydrological forecasts varied according to the configuration of the forecasting system, the forecast lead time, and the catchment size. The combination of the precipitation post-processor and the quantification of uncertainty from initial conditions showed the best results. When all sources of uncertainty were quantified, the contribution of the precipitation post-processor to provide better streamflow forecasts was not remarkable and, in some cases, it even deteriorated the overall performance of the hydrometeorological forecasting system. Our study provides an in-depth investigation on how improvements brought by a precipitation post-processor to the quality of the inputs to a hydrological forecasting model can be cancelled along the forecasting chain, depending on how the hydrometeorological forecasting system is configured and on how the other sources of hydrological forecasting uncertainty (initial conditions and model structure) are considered and accounted for. This has implications for the choices users might make when designing new or enhancing existing hydrometeorological ensemble forecasting systems.

Possibilities of Using Neuro-Fuzzy Models for Post-Processing of Hydrological Forecasts

When issuing hydrological forecasts and warnings for individual profiles, the aim is to achieve the best possible results. Hydrological forecasts themselves are burdened by an error (uncertainty) at the inputs (precipitation forecast) as well as on the side of the hydrological model used. The aim of the method described in this article is to reduce the error of the hydrological model using post-processing the model results. Models based on neuro-fuzzy models were selected for the post-processing itself. The whole method was tested on 12 profiles in the Czech Republic. The catchment size of the individual profiles ranged from 90 to 4500 km2 and the profiles varied in their character, both in terms of elevation as well as land cover. After finding the suitable model architecture and introducing supporting algorithms, there was an improvement in the results for the individual profiles for selected criteria by on average 5–60% (relative culmination error, mean square error) compared to the results of re-simulation of the hydrological model. The results of the application show that the method was able to improve the accuracy of hydrological forecasts and thus could contribute to better management of flood situations.

Wastewater Surveillance of SARS-CoV-2 across 40 U.S. states

Wastewater-based disease surveillance is a promising approach for monitoring community outbreaks. Here we describe a nationwide campaign to monitor SARS-CoV-2 in the wastewater of 159 counties in 40 U.S. states, covering 13% of the U.S. population from February 18 to June 2, 2020. Out of 1,751 total samples analyzed, 846 samples were positive for SARS-CoV-2 RNA, with overall viral concentrations declining from April to May. Wastewater viral titers were consistent with, and appeared to precede, clinical COVID-19 surveillance indicators, including daily new cases. Wastewater surveillance had a high detection rate (>80%) of SARS-CoV-2 when the daily incidence exceeded 13 per 100,000 people. Detection rates were positively associated with wastewater treatment plant catchment size. To our knowledge, this work represents the largest-scale wastewater-based SARS-CoV-2 monitoring campaign to date, encompassing a wide diversity of wastewater treatment facilities and geographic locations. Our findings demonstrate that a national wastewater-based approach to disease surveillance may be feasible and effective.

Effects of climate change and human interactions on water balance dynamics in the River Vistula basin

&lt;div&gt; &lt;p&gt;Water balance modelling is often applied in studies of climate and human impacts on water resources. Annual water balance is usually derived based on precipitation, discharge and temperature observations under an assumption of negligible changes in annual water storage in a catchment. However, that assumption might be violated during very dry or very wet years. In this study we apply groundwater level measurements to improve water balance modelling in nine sub-catchments of the River Vistula basin starting from the river sources downstream. Annual and inter-annual water balance is studied using a Budyko framework to assess actual evapotranspiration and total water supply. We apply the concept of effective precipitation to account for possible losses due to water interception by vegetation. Generalised Likelihood Uncertainty Estimation GLUE is used to account for parameter and structural model uncertainty, together with the application of eight Budyko-type equations. Seasonal water balance models show large errors for winter seasons while summer and annual water balance models follow the Budyko framework. The dryness index is much smaller in winter than in summer for all sub-catchments. The spatial variability of water balance modelling errors indicate an increasing uncertainty of model predictions with an increase in catchment size. The results show that the added information on storage changes in the catchments provided by groundwater level observations largely improves model accuracy. The results also indicate the need to model groundwater level variability depending on external factors such as precipitation and evapotranspiration and human interventions. The modelling tools developed will be used to assess future water balance in the River Vistula basin under different water management scenarios and climate variability.&lt;/p&gt; &lt;/div&gt;