Post-processing ECMWF precipitation and temperature ensemble reforecasts for operational hydrologic forecasting at various spatial scales

Verkade, Jan; Brown, James Dean; Reggiani, P.; Weerts, Albrecht

doi:10.1016/j.jhydrol.2013.07.039

Cited by 159 publications

(156 citation statements)

References 50 publications

Supporting

Mentioning

149

Contrasting

Order By: Relevance

“…This has also been observed by other authors. For instance, Verkade et al (2013) found that post-processing does not improve on all qualities at all lead times and at all levels of the verifying observations. The cause is rooted in the impossibility of the post-processing approaches to replace adequate model representation of physical processes (Haerter et al, 2011).…”

Section: Evaluating Precipitation Forecasts From the Wrf Modelmentioning

confidence: 96%

Streamflow forecasts from WRF precipitation for flood early warning in mountain tropical areas

Rogelis

Werner

2018

Hydrol. Earth Syst. Sci.

View full text Add to dashboard Cite

Abstract. Numerical weather prediction (NWP) models are fundamental to extend forecast lead times beyond the concentration time of a watershed. Particularly for flash flood forecasting in tropical mountainous watersheds, forecast precipitation is required to provide timely warnings. This paper aims to assess the potential of NWP for flood early warning purposes, and the possible improvement that bias correction can provide, in a tropical mountainous area. The paper focuses on the comparison of streamflows obtained from the post-processed precipitation forecasts, particularly the comparison of ensemble forecasts and their potential in providing skilful flood forecasts. The Weather Research and Forecasting (WRF) model is used to produce precipitation forecasts that are post-processed and used to drive a hydrologic model. Discharge forecasts obtained from the hydrological model are used to assess the skill of the WRF model. The results show that post-processed WRF precipitation adds value to the flood early warning system when compared to zeroprecipitation forecasts, although the precipitation forecast used in this analysis showed little added value when compared to climatology. However, the reduction of biases obtained from the post-processed ensembles show the potential of this method and model to provide usable precipitation forecasts in tropical mountainous watersheds. The need for more detailed evaluation of the WRF model in the study area is highlighted, particularly the identification of the most suitable parameterisation, due to the inability of the model to adequately represent the convective precipitation found in the study area.

show abstract

Section: Evaluating Precipitation Forecasts From the Wrf Modelmentioning

confidence: 96%

Streamflow forecasts from WRF precipitation for flood early warning in mountain tropical areas

Rogelis

Werner

2018

Hydrol. Earth Syst. Sci.

View full text Add to dashboard Cite

show abstract

“…Typically, these techniques relate a combination of model inputs and/or outputs to the model error distribution. Various post-processors have been developed and applied to hydrologic modelling, such as a meta-Gaussian error model (Montanari and Brath, 2004), UNEEC (Solomatine and Shrestha, 2009), quantile regression (Weerts et al, 2011), and DUMBRAE (Pianosi and Raso, 2012). Quantile regression (QR) is a relatively straightforward post-processing technique that relates the probability of residual errors to the model forecast (the predictand) by a regression model that is derived from historical forecasts and observations.…”

mentioning

confidence: 99%

“…Quantile regression (QR) is a relatively straightforward post-processing technique that relates the probability of residual errors to the model forecast (the predictand) by a regression model that is derived from historical forecasts and observations. QR has been successfully applied for uncertainty quantification in hydrologic forecasts with various modifications (Weerts et al, 2011;Verkade et al, 2013;Roscoe et al, 2012;López López et al, 2014;Hoss and Fischbeck, 2015), whereas UNEEC involves a machine learning technique for building a non-linear regression model of error quantiles (Solomatine and Shrestha, 2009). UNEEC includes three steps: (1) fuzzy clustering of input data in the space of "relevant" variables; (2) estimating the probability distribution function of residual errors for each cluster and (3) building a machine learning model (e.g.…”

mentioning

confidence: 99%

Residual uncertainty estimation using instance-based learning with applications to hydrologic forecasting

Wani

Beckers

Weerts

et al. 2017

Hydrol. Earth Syst. Sci.

View full text Add to dashboard Cite

Abstract. A non-parametric method is applied to quantify residual uncertainty in hydrologic streamflow forecasting. This method acts as a post-processor on deterministic model forecasts and generates a residual uncertainty distribution. Based on instance-based learning, it uses a k nearestneighbour search for similar historical hydrometeorological conditions to determine uncertainty intervals from a set of historical errors, i.e. discrepancies between past forecast and observation. The performance of this method is assessed using test cases of hydrologic forecasting in two UK rivers: the Severn and Brue. Forecasts in retrospect were made and their uncertainties were estimated using kNN resampling and two alternative uncertainty estimators: quantile regression (QR) and uncertainty estimation based on local errors and clustering (UNEEC). Results show that kNN uncertainty estimation produces accurate and narrow uncertainty intervals with good probability coverage. Analysis also shows that the performance of this technique depends on the choice of search space. Nevertheless, the accuracy and reliability of uncertainty intervals generated using kNN resampling are at least comparable to those produced by QR and UNEEC. It is concluded that kNN uncertainty estimation is an interesting alternative to other post-processors, like QR and UNEEC, for estimating forecast uncertainty. Apart from its concept being simple and well understood, an advantage of this method is that it is relatively easy to implement.

show abstract

“…A couple of studies have quantified the effects on streamflow skill by preprocessing either seasonal (Crochemore et al, 2016) or medium range forecasts (Verkade et al, 2013) forecasts. Other studies have assessed the efficiency of postprocessing streamflow forecasts only (Bogner et al, 2016;Madadgar et al, 2014;Ye et al, 2015;Zhao et al, 15 2011;Wood and Schaake, 2008).…”

mentioning

confidence: 99%

“…To our knowledge, only Roulin and Vannitsem, (2015); Yuan and Wood, (2012) and Zalachori et al, (2012) have compared the additional gain in skill of doing both pre-and postprocessing. The previous studies have shown that improvements made by preprocessing the forcings do not necessarily translate into improvements in streamflow forecasts (Verkade et al, 2013;Zalachori et al, 2012). Improvements are larger when postprocessing is done, and a combination of pre-and postprocessing provides the best results (Yuan and Wood, 2012;Zalachori et al, 2012).…”

mentioning

confidence: 99%

Seasonal streamflow forecasts in the Ahlergaarde catchment, Denmark: effect of preprocessing and postprocessing on skill and statistical consistency

Lucatero¹,

Madsen²,

Refsgaard³

et al. 2017

Preprint

View full text Add to dashboard Cite

Abstract.In the present study we analyze the effect of bias adjustments in both meteorological and streamflow forecasts on skill and reliability of monthly average streamflow and low flow forecasts. Both raw and pre-processed meteorological 10 seasonal forecast from the European Center for Medium-Range Weather Forecasts (ECMWF) are used as inputs to a spatially distributed, coupled surface -subsurface hydrological model based on the MIKE SHE code in order to generate streamflow predictions up to seven months in advance. In addition to this, we postprocess streamflow predictions using an empirical quantile mapping that adjusts the predictive distribution in order to match the observed one. Bias, skill and statistical consistency are the qualities evaluated throughout the forecast generating strategies and we analyze where the 15 different strategies fall short to improve them. ECMWF System 4-based streamflow forecasts tend to show a lower accuracy level than those generated with an ensemble of historical observations, a method commonly known as Ensemble Streamflow Prediction (ESP). This is particularly true at longer lead times, for the dry season and for streamflow stations that exhibit low hydrological model errors. Biases in the mean are better removed by postprocessing that in turn is reflected in the higher level of statistical consistency. However, in general, the reduction of these biases is not enough to ensure a higher level of 20 accuracy than the ESP forecasts. This is true for both monthly mean and minimum yearly streamflow forecasts. We highlight the importance of including a better estimation of the initial state of the catchment, which will increase the capability of the system to forecast streamflow at longer leads.

show abstract

Post-processing ECMWF precipitation and temperature ensemble reforecasts for operational hydrologic forecasting at various spatial scales

Cited by 159 publications

References 50 publications

Streamflow forecasts from WRF precipitation for flood early warning in mountain tropical areas

Streamflow forecasts from WRF precipitation for flood early warning in mountain tropical areas

Residual uncertainty estimation using instance-based learning with applications to hydrologic forecasting

Seasonal streamflow forecasts in the Ahlergaarde catchment, Denmark: effect of preprocessing and postprocessing on skill and statistical consistency

Contact Info

Product

Resources

About