Figure 3 - uploaded by Shaun Harrigan
Content may be subject to copyright.
UK-wide mean ESP skill scores across all 314 catchments and 12 forecast initialisation months for all 365 lead times (LTs) for both MSESS (blue line) and CRPSS (red line) skill metrics. The range of skill scores across catchments at each LT is shown by semitransparent 5th and 95th percentile bands. Vertical lines represent eight commonly used operational forecasting LTs from short (days) to annual (12-months).
Source publication
Skilful hydrological forecasts at sub-seasonal to seasonal lead times would
be extremely beneficial for decision-making in water resources management,
hydropower operations, and agriculture, especially during drought conditions.
Ensemble streamflow prediction (ESP) is a well-established method for
generating an ensemble of streamflow forecasts in t...
Contexts in source publication
Context 1
... mean ESP skill across all catchments and initialisation months decays exponentially as a function of lead time for both the MSESS and CRPSS metrics (Fig. 3 Fig. 2a with Fig. 2d, as discussed further in Sect. ...
Context 2
... decay in skill with LT as shown in Fig. 3 also occurs across all initialisation months (Figs. 4 and 5). Whilst mean ESP skill tends towards zero for longer LTs, there are many catchments with much higher skill scores than average. For example, at a 1-month LT initialised in July the average UK-wide ESP skill is moderate (MSESS = 0.303), but 42 catchments have high skill (MSESS ...
Similar publications
Skilful hydrological forecasts at sub-seasonal to seasonal lead times would be extremely beneficial for decision-making in water resources management, hy-dropower operations, and agriculture, especially during drought conditions. Ensemble streamflow prediction (ESP) is a well-established method for generating an ensemble of streamflow forecasts in...
Citations
... As with the SRF method, there is no formal statement of the confidence of any of the forecasts over any duration. However, Harrigan et al. (2017) have investigated the skill of the ESP method in the UK, but use a different rainfall-runoff model to the one used by the UKHO for the period covered in the assessment presented in this study. ...
The UK Hydrological Outlook (UKHO) is a seasonal forecast of future river flows and groundwater levels. The UKHO contains both presentations of outputs from models simulating future conditions and a high-level summary. The summary is produced by an expert panel of forecasters that considers the model outputs together with other recent hydrological and meteorological information. Whilst the skill and uncertainty of the individual models have been explored and published, this study sets out to establish the performance of the high-level summary, and presents such an assessment of the river flow forecasts at the 1-month timescale. Both qualitative and quantitative assessments are presented and compared with two naïve forecasting methods. The UKHO summary is found to have a similar Gerrity skill score to a ‘same as last month’ forecast, an outcome that generates suggestions for improvements in how the different model outputs should be considered and presented in the high-level summary.
... In ESP, hydrologic models are forced with historical sequence of climate data, such as precipitation, temperature and/or potential evapotranspiration during the time of forecast, providing a plausible range of future streamflow states [19]. The method assumes that the forcing data and model are perfect, that is, there are no errors in the initial hydrological conditions (IHCs) [20]. ...
Hydrologic models are an approximation of reality, and thus, are not able to perfectly simulate observed streamflow because of various sources of uncertainty. On the other hand, skillful operational hydrologic forecasts are vital in water resources engineering and management for preparedness against flooding and extreme events. Multi-model techniques can be used to help represent and quantify various uncertainties in forecasting. In this paper, we assess the performance of a Multi-model Seasonal Ensemble Streamflow Prediction (MSESP) scheme coupled with statistical post-processing techniques to issue operational uncertainty for the Manitoba Hydrologic Forecasting Centre (HFC). The Ensemble Streamflow Predictions (ESPs) from WATFLOOD and SWAT hydrologic models were used along with four statistical post-processing techniques: Linear Regression (LR), Quantile Mapping (QM), Quantile Model Averaging (QMA), and Bayesian Model Averaging (BMA)]. The quality of MSESP was investigated from April to July with a lead time of three months for the Upper Assiniboine River Basin (UARB) at Kamsack, Canada. While multi-model ESPs coupled with post-processing techniques improve predictability (in general), results suggest that additional avenues for improving the skill and value of seasonal streamflow prediction. Next steps towards an operational ESP system include adding more operationally used models, improving models calibration methods to reduce model bias, increasing ESP sample size, and testing ESP schemes at multiple lead times, which, once developed, will not only help HFCs in Canada but would also help Centers South of the Border.
... In this approach, observed historical meteorological conditions are used to generate the streamflow forecasts. Several studies have shown that Ensemble Streamflow prediction (ESP) can provide skillful seasonal streamflow prediction (Harrigan et al., 2017) and seasonal hydrometeorological prediction systems have difficulties to outperform the ESP beyond one month lead time (Arnal et al., 2018;Lucatero et al., 2018). Irrespective of the methodology used, the performance of ensemble streamflow forecasts depend to a large degree on 30 the initial condition within the catchment especially for small catchments (Van Dijk et al., 2013;Thirel et al., 2010). ...
... Many studies have shown the potential of this method to provide skillful streamflow predictions at the subseasonal to seasonal time scale in Europe (e.g. Arnal et al., 2018;Harrigan et al., 2017). The basic principle behind the ESP is to create an ensemble of streamflows based on known initial conditions and forced by historic climate sequences. ...
... Furthermore, the substantial decrease in skill within the first days suggest that the ESP predictions are not able to capture the exact evolution but can rather be used to estimate the general behavior within the upcoming weeks. This is in agreement with 15 the exponential decay in skill with increasing lead time found for ESP forecasts in UK catchments (Harrigan et al., 2017). If the NWP predictions are used to predict the streamflows, the skill can clearly be enhanced but in most cases only if the driving meteorological predictions are pre-processed. ...
Traditional Ensemble Streamflow Prediction systems (ESP) are known to provide a valuable baseline to predict streamflows at the subseasonal to seasonal timescale. They exploit a combination of initial conditions and past meteorological observations, and can often provide useful forecasts of the expected streamflow in the upcoming month. In recent years, numerical weather prediction (NWP) models for subseasonal to seasonal timescales have made large progress and can provide added value to such a traditional ESP approach. Prior of using such meteorological predictions two major problems need to be solved: the correction of biases, and downscaling to account to increase the spatial resolution. Various methods exist to overcome these problems, but the potential of using NWP information and the relative merit of the different statistical and modeling steps remains open. To address this question, we compare a traditional ESP system with a subseasonal hydrometeorological ensemble prediction system in three alpine catchments with varying hydroclimatic conditions with areas between 80 and 1700km². Uncorrected and corrected (pre-processed) temperature and precipitation reforecasts from the ECMWF subseasonal NWP model are used to run the hydrological simulations and the performance of the resulting streamflow predictions is assessed with commonly used verification scores characterizing different aspects of the forecasts (ensemble mean and spread). Our results indicate that the NWP based approach can provide superior prediction than the ESP approach, especially at shorter lead times. In snow-dominated catchments the pre-processing of the meteorological input further improves the performance of the predictions. This is most pronounced in late winter and spring when snow melting occurs. Moreover, our results highlight the importance of snow related processes for subseasonal streamflow predictions in mountainous regions.
... At lead times of months to a few seasons, it is common for initial watershed moisture and to a lesser extent energy (collectively, IHCs) to exert a clear if not dominant influence on hydrologic forecastsa result of the inertia and persistence found in hydrologic systems. Where groundwater contributes a significant fraction of river flow, IHCs can even affect forecast outcomes for lead times of over a year (Harrigan et al. 2017). ...
In many parts of the world, water resources systems manage sub-seasonal to
seasonal (S2S) variability in climate and runoff in part through the use of operational
streamflow forecasts, supplemented by predictions of climate and other hydrologic
variables. S2S hydrologic forecasts are produced through both statistical and
dynamical (model-based) approaches, and separate S2S forecasts may be combined
in multi-model frameworks to increase their skill. Statistical post-processing can be
used to enhance the skill and reliability of model-based S2S predictions, and to
reduce bias, as well as to merge forecasts from multiple approaches. This chapter
describes seasonal hydrologic forecast approaches and products, and presents
common techniques used in both the post-processing of single ensemble forecast
series as well as the combination of multiple forecasts. Also discussed are the
sources of S2S hydrological predictability and particular challenges and opportunities
related to post-processing seasonal hydrologic predictions, for which the sample
sizes of past simulations, observations and predictions are relatively more limited
than in the context of short to medium range prediction.
... The ESP method has been shown to be skilful in the UK, with average mean-squared-error skill scores across the year of 0.3, 0.17 and 0.11 for 1-, 3-and 6month lead times, respectively (Harrigan et al. 2017), for a diverse sample of 314 catchments. Skill varies depending on forecast initialization month and is generally higher for seasonal forecasts initialized in wetter months than for dry. ...
... Another difficulty, inherent to any operational system, is the day-to-day operational running of methods in parallel with scientific development and evaluation of new tools. While skill assessment of each of the methods has been undertaken (Svensson 2016, Mackay et al. 2015, Bell et al. 2017, Harrigan et al. 2017, there remains an ongoing need for research to fully understand the performance of the methods, and how this can be better incorporated into the final Hydrological Outlook product. Furthermore, work is underway aimed at evaluating the reliability of the summary page, in a multi-method skill assessment framework, and establishing how best to incorporate the knowledge gained during the skill evaluation into operational decision making. ...
This paper describes the development of the first operational seasonal hydrological forecasting service for the UK, the Hydrological Outlook UK (HOUK). Since June 2013, this service has delivered monthly forecasts of streamflow and groundwater levels, with an emphasis on forecasting hydrological conditions over the next three months, accompanied by outlooks over longer time horizons. This system is based on three complementary approaches combined to produce the outlooks: (i) national-scale modelling of streamflow and groundwater levels based on dynamic seasonal rainfall forecasts, (ii) catchment-scale modelling where streamflow and groundwater level models are driven by historical meteorological forcings (i.e. the Ensemble Streamflow Prediction, ESP, approach), and (iii) a catchment-scale statistical method based on persistence and historical analogues. This paper provides the background to the Hydrological Outlook, describes the various component methods in detail and then considers the impact and usefulness of the product. As an example of a multi-method, operational seasonal hydrological forecasting system, it is hoped that this overview provides useful information and context for other forecasting initiatives around the world.
Seasonal streamflow forecasts facilitate water allocation, reservoir operation, flood risk management, and crop forecasting. They are generally computed by forcing hydrological models with outputs from general circulation models (GCMs) or using large-scale climate indices as predictors in statistical models. In contrast, hybrid statistical-dynamical forecasts (combining statistical methods with dynamical
climate predictions) are still uncommon, and their skill is largely unknown. Here we conduct systematic forecasting of seasonal streamflow using eight GCMs from the North American Multi-Model Ensemble,
0.5–9.5 months ahead, at 290 stream gauges in the U.S. Midwest. Probabilistic forecasts are developed for low to high streamflow using predictors that reflect climatic and anthropogenic influences. Results indicate thatGCM forecasts of climate and antecedent climatic conditions enhance seasonal streamflow predictability; while land cover and population density predictors decrease biases or enhance skill in certain catchments. This paper paves the way for novel forecasting approaches using dynamical GCM predictions within statistical frameworks.
