Downscaling of surface moisture flux and precipitation in the Ebro Valley (Spain) using analogues and analogues followed by random forests and multiple linear regression

Ibarra-Berastegi, Gabriel; Sáenz, Jon; Ezcurra, A.; Ηλίας, Α.; Argandoña, Javier Díaz de; Errasti, Iñigo

doi:10.5194/hess-15-1895-2011

Cited by 41 publications

(31 citation statements)

References 49 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Analogue methods have been applied in different regions of the world with very diverse climates, e.g. Switzerland , Australia (Timbal and McAvaney, 2001), central Sweden (Wetterhall et al, 2005), Punjab (India) (Raje and Mujumdar, 2011), southeast USA (Zhang and Georgakakos, 2012), the Alpine region (Themeßl et al, 2011), and northeast Spain (Ibarra-Berastegi et al, 2011).…”

Section: Statistical Downscaling Methodsmentioning

confidence: 99%

Optimising predictor domains for spatially coherent precipitation downscaling

Radanovics¹,

Vidal²,

Sauquet³

et al. 2013

Hydrol. Earth Syst. Sci.

View full text Add to dashboard Cite

Abstract. Statistical downscaling is widely used to overcome the scale gap between predictors from numerical weather prediction models or global circulation models and predictands like local precipitation, required for example for medium-term operational forecasts or climate change impact studies. The predictors are considered over a given spatial domain which is rarely optimised with respect to the target predictand location. In this study, an extended version of the growing rectangular domain algorithm is proposed to provide an ensemble of near-optimum predictor domains for a statistical downscaling method. This algorithm is applied to find five-member ensembles of near-optimum geopotential predictor domains for an analogue downscaling method for 608 individual target zones covering France. Results first show that very similar downscaling performances based on the continuous ranked probability score (CRPS) can be achieved by different predictor domains for any specific target zone, demonstrating the need for considering alternative domains in this context of high equifinality. A second result is the large diversity of optimised predictor domains over the country that questions the commonly made hypothesis of a common predictor domain for large areas. The domain centres are mainly distributed following the geographical location of the target location, but there are apparent differences between the windward and the lee side of mountain ridges. Moreover, domains for target zones located in southeastern France are centred more east and south than the ones for target locations on the same longitude. The size of the optimised domains tends to be larger in the southeastern part of the country, while domains with a very small meridional extent can be found in an east-west band around 47 • N. Sensitivity experiments finally show that results are rather insensitive to the starting point of the optimisation algorithm except for zones located in the transition area north of this east-west band. Results also appear generally robust with respect to the archive length considered for the analogue method, except for zones with high interannual variability like in the Cévennes area. This study paves the way for defining regions with homogeneous geopotential predictor domains for precipitation downscaling over France, and therefore de facto ensuring the spatial coherence required for hydrological applications.

show abstract

Section: Statistical Downscaling Methodsmentioning

confidence: 99%

Optimising predictor domains for spatially coherent precipitation downscaling

Radanovics¹,

Vidal²,

Sauquet³

et al. 2013

Hydrol. Earth Syst. Sci.

View full text Add to dashboard Cite

show abstract

“…In order to achieve a rigorous evaluation of the performance of statistical models, the original database was split into two datasets, one to fit and train the models (2007)(2008)(2009)(2010) and the other to test the models on independent data (2011-2014). RF uses train and test datasets [18,55]; this method does not require validation datasets, such as the neural network method. The number of cases is shown in Table 1.…”

Section: Methodsmentioning

confidence: 99%

Sensitivity Studies for a Hybrid Numerical–Statistical Short-Term Wind and Gust Forecast at Three Locations in the Basque Country (Spain)

et al. 2019

View full text Add to dashboard Cite

This study evaluates the performance of statistical models applied to the output of numerical models for short-term (1-24 h) hourly wind forecasts at three locations in the Basque Country. The target variables are horizontal wind components and the maximum wind gust at 3 h intervals. Statistical approaches such as persistence, analogues, linear regression, and random forest (RF) are used. The verification statistics used are coefficient of determination (R 2 ) and root mean square error (RMSE). Statistical models use three inputs: (1) Local wind observations; (2) extended EOFs (empirical orthogonal functions) derived from past local observations and ERA-Interim variables in a previous 24-h period covering a domain around the area of study; and (3) wind forecasts provided by ERA-Interim. Results indicate that, for horizons less than 1-4 h, persistence is the best model. For longer predictions, RF provides the best forecasts. For horizontal components at 4-24 h horizons, RF slightly outperformed ERA-Interim wind forecasts. For gust, RF performs better than ERA-Interim for all the horizons. Persistence is the most influential factor for 2-5 h. Beyond this horizon, predictors from the ERA-Interim wind forecasts led the contribution. Hybrid numerical-statistical methods can be used to improve short-term wind forecasts.Atmosphere 2020, 11, 45 2 of 22Wind predictions are issued based on meteorological numerical weather forecast (NWF) systems, in which data assimilation (DA) plays an important role in order to properly correct errors in the forecasting model by the use of observations in the data assimilation stages. In these NWF systems, the data assimilation methods optimally estimate the state of the atmosphere [19,20] and a numerical weather forecast model is used to predict the future states of the atmosphere. Two examples of these advanced NWP systems are the ERA-Interim [21] global atmospheric reanalysis model, from 1979 until 2019 [21], and NCEP re-analyses [22]. They are not real-time operational NWF systems, since the DA and the forecasting models are frozen in time, but they are re-run for the whole period using all the available observations commonly used in numerical weather forecasts in order to avoid, as much as possible, errors in the record due to changes in the structure (architecture, resolution, kind of observations processed) of the archives derived from operational NWF systems. In the case of the ERA-Interim re-analyses, besides analyses, forecasts are also provided starting at the analysis times. These forecasts are the ones used in this paper as a surrogate of an operational model. However, re-analyses are quite coarse-resolution models when compared with operational models of the same generation. In the particular case of the ERA-Interim, the spatial resolution is 0.75 • , too coarse for local wind forecasts.To fill this resolution gap, numerical and statistical downscaling techniques are customarily used. Numerical downscaling increases the resolution by nesting a higher-resolution model, such ...

show abstract

“…This method was initially used in the field of weather forecasting (Lorenz, 1969) and has been introduced by Zorita et al (1995) for downscaling purposes. The analog method as a downscaling approach has been extensively applied to downscale climate change scenarios (Imbert and Benestad, 2005;Timbal et al, 2003), precipitation (Fernández-Ferrero et al, 2010Hidalgo et al, 2008;Ibarra-Berastegi et al, 2011;Matulla et al, 2008;Maurer and Hidalgo, 2008;Maurer et al, 2010;Wetterhall et al, 2005), air temperature Maurer and Hidalgo, 2008;Maurer et al, 2010;Timbal and McAvaney, 2001), surface moisture flux (Ibarra-Berastegi et al, 2011), wildfire (Abatzoglou and Brown, 2012), and daily ET o forecasts (Tian and Martinez, in press).…”

Section: Introductionmentioning

confidence: 99%