References of "Doutreloup, Sébastien"
     in
Bookmark and Share    
Full Text
Peer Reviewed
See detailComparaison entre le profil vertical de la vitesse du vent observé dans les basses couches de la troposphère et celui simulé par le modèle WRF en Belgique
Doutreloup, Sébastien ULg; Fettweis, Xavier ULg; Beaumet, Julien ULg et al

in Camberlin, Pierre; Richard, Yves (Eds.) Actes du XXVIIe Colloque de l'Association Internationale de Climatologie : CLIMAT : SYSTÈME & INTERACTIONS (2014, July 02)

In the framework of FLEXIPAC project funded by the "RELIABLE" program of Walloon Region (Belgium), the Laboratory of Climatology Topoclimatology (LCT) of the University of Liège (Belgium) aims to adjust ... [more ▼]

In the framework of FLEXIPAC project funded by the "RELIABLE" program of Walloon Region (Belgium), the Laboratory of Climatology Topoclimatology (LCT) of the University of Liège (Belgium) aims to adjust the WRF regional model (v.3.4.) forced by the ERA-Interim reanalysis for Belgium. Our analysis shows that wind speeds at 100m simulated by WRF are systematically overestimated compared to wind speeds extracted from wind productions of two wind farms. In order to solve this problem, four ways are considered in this contribution. The first way is to compare the WRF model with the reanalysis data. The second way is to test the influence of the spatial resolution by running WRF with a finer resolution. The third way is to smooth WRF outputs, where in order to analyze the variability created by the model. And finally, the fourth way is to compare the WRF model with the MAR (v3.3.) regional model. This last way seems to confirm that the MAR model better simulates wind speeds at 10m and at 100m than the WRF model. [less ▲]

Detailed reference viewed: 18 (1 ULg)
Full Text
Peer Reviewed
See detailComparison between wind speed observed at 100m height and wind speeds simulated by the WRF and MAR models
Doutreloup, Sébastien ULg; Fettweis, Xavier ULg; Erpicum, Michel ULg et al

Poster (2014, June 06)

In the context of FLEXIPAC project funded by "RELIABLE" program of Walloon Region (Belgium), the Laboratory of Climatology and Topoclimatology (LCT) of the University of Liège (Belgium) aims to adjust the ... [more ▼]

In the context of FLEXIPAC project funded by "RELIABLE" program of Walloon Region (Belgium), the Laboratory of Climatology and Topoclimatology (LCT) of the University of Liège (Belgium) aims to adjust the WRF regional model (v.3.4.) forced by ERA-Interim model. Our analysis shows that wind speeds at 100m height simulated by WRF are systematically overestimated compared to wind speeds extracted from wind productions of two wind farms. In order to identify this problem, four comparisons were performed in this contribution. Firstly, we compare WRF model with reanalysis based forcing model. Secondly, we compare two WRF simulations, where one of them has a more precise spatial resolution. Thirdly, we smooth WRF outputs in time (6-hr running mean) in order to study the accuracy of the 30-min variability generated by WRF model. Finally, we comp compare the WRF model with the MAR (v3.3.) regional model using the same forcing at its lateral boundaries. This last one seems to suggest that the MAR model better simulates wind speeds at 10m and at 100m than WRF model and then that wind speed underestimation by WRF is well linked to the WRF physics itself. [less ▲]

Detailed reference viewed: 41 (4 ULg)
Full Text
Peer Reviewed
See detailRECOURS À UNE CAMPAGNE DE MESURES TOPOCLIMATIQUES APPLIQUÉE POUR UNE ÉTUDE COMPARATIVE DE VARIABLES MÉTÉOROLOGIQUES CONCOMITANTES PROVENANT DES STATIONS MÉTÉOROLOGIQUES DE BRUGGE ET DU MONT RIGI EN BELGIQUE
Pirard, Xavier ULg; Jorion, Nicolas ULg; Doutreloup, Sébastien ULg et al

Poster (2012, September 07)

Une station météorologique automatique munie d’un mât de 10 mètres a été installée à l’Ouest de l’agglomération urbaine de Brugge (Belgique) par le Laboratoire de Climatologie et Topoclimatologie de l’ULg ... [more ▼]

Une station météorologique automatique munie d’un mât de 10 mètres a été installée à l’Ouest de l’agglomération urbaine de Brugge (Belgique) par le Laboratoire de Climatologie et Topoclimatologie de l’ULg. Elle a été équipée afin de confirmer la bonne qualité des prévisions météorologiques du modèle WRF établies dans le cadre du projet européen TWENTIES. Les données récoltées de minute en minute par cette station météorologique offrent également l’opportunité de mettre en évidence le détail de situations météorologiques bien particulières comme celles correspondant aux passages de fronts. [less ▲]

Detailed reference viewed: 44 (6 ULg)
Full Text
Peer Reviewed
See detailÉVOLUTION DU COMPORTEMENT DU VENT ET DE SON POTENTIEL POUR LA PRODUCTION D'ÉNERGIE ÉOLIENNE DURANT LES 30 DERNIÈRES ANNÉES : LE CAS DE LA BELGIQUE
Doutreloup, Sébastien ULg; Fettweis, Xavier ULg; Erpicum, Michel ULg

in Bigot, Sylvain; Rome, Sandra (Eds.) Les climats régionaux : observation et modélisation. (Actes du colloque organisé à Grenoble du mercredi 5 au samedi 8 septembre 2012) (2012, September)

Chaque année, le nombre d'éoliennes dans le monde augmente de façon significative suite notamment aux politiques encourageant les productions d'énergie verte afin d’atténuer le réchauffement climatique ... [more ▼]

Chaque année, le nombre d'éoliennes dans le monde augmente de façon significative suite notamment aux politiques encourageant les productions d'énergie verte afin d’atténuer le réchauffement climatique. Toutefois, ce type d'énergie est tributaire de la météo. Cela implique que la production d'énergie éolienne est irrégulière à courte échelle de temps. Cependant, la disponibilité d’électricité de courtes périodes de temps est très importante à connaitre pour les producteurs d'énergie ainsi que pour les gestionnaires de réseaux. Pour ces raisons, il nous parait primordial d’analyser l’évolution de l’intermittence de la vitesse du vent sur les 30 dernières années (1979-2009). Pour ce faire nous utilisons le modèle WRF forcé par les réanalyses ERA-Interim, les réanalyses NCEP2 et certains modèles du GIEC (base de données CMIP5). [less ▲]

Detailed reference viewed: 87 (5 ULg)
Full Text
See detailEvaluation of the MAR and WRF regional climate models over Svalbard
Lang, Charlotte ULg; Fettweis, Xavier ULg; Doutreloup, Sébastien ULg et al

Conference (2012, June 01)

It is well known that high latitude zones are very sensitive to climate change. As a result of global warming, ice sheet melting has increased which in turn has an influence on climate through ... [more ▼]

It is well known that high latitude zones are very sensitive to climate change. As a result of global warming, ice sheet melting has increased which in turn has an influence on climate through modifications of the thermohaline circulation, feedback of ice albedo, sea level rise... Svalbard is an archipelago between 74 and 81°lat N and 60 percent of its area (62 248 km2) is covered with glaciers and ice sheets. The impact of global warming on the Svalbard cryosphere can be estimated with climate models. However, we need to use regional climate models as they offer the possibility of a higher resolution than general circulation models. We have ran two regional climate models (MAR and WRF) at a 10-kilometre resolution between 2006 and 2010 over Svalbard and compared their simulated climate to near surface measurements at several weather stations through the archipelago in order to determine which one of them could best represent the Svalbard climate. [less ▲]

Detailed reference viewed: 42 (2 ULg)
Full Text
See detailEvaluation of the regional climate model WRF over Svalbard
Lang, Charlotte ULg; Fettweis, Xavier ULg; Doutreloup, Sébastien ULg et al

Poster (2012, April 24)

It is well known that high latitude zones are very sensitive to climate change. As a result of global warming, ice sheet melting has increased which in turn has an influence on climate through ... [more ▼]

It is well known that high latitude zones are very sensitive to climate change. As a result of global warming, ice sheet melting has increased which in turn has an influence on climate through modifications of the thermohaline circulation, feedback of ice albedo, sea level rise, … Svalbard is an archipelago between 74 and 81°lat N and 60 percent of its area (62 248 km2) is covered with glaciers and ice sheets. The impact of global warming on the Svalbard cryosphere can be estimated with climate models. However, we need to use regional climate models as they offer the possibility of a higher resolution than general circulation models. We have ran two regional climate models (MAR and WRF) at a 10-kilometre resolution between 2006 and 2010 over Svalbard and compared their simulated climate to near surface measurements at several weather stations through the archipelago in order to determine which one of them could best represent the Svalbard climate. [less ▲]

Detailed reference viewed: 52 (5 ULg)
Full Text
See detailEstimation, prévision et contrôle du gisement solaire en région wallonne
Doutreloup, Sébastien ULg; Fettweis, Xavier ULg; Erpicum, Michel ULg

Poster (2012, April 23)

Présentation des activités et des recherches du Laboratoire de Climatologie et Topoclimatologie de l'Université de Liège dans le domaine du rayonnement solaire

Detailed reference viewed: 40 (1 ULg)
Full Text
Peer Reviewed
See detailEvolution of wind behaviour and of its potential for wind power production in Belgium during the last 22 years : a comparison between WRF forced by NCEP2 reanalysis and WRF forced by ERA-INTERIM reanalysis
Doutreloup, Sébastien ULg; Fettweis, Xavier ULg; Erpicum, Michel ULg

Conference (2011, September 16)

The number of wind turbines in the world grows significantly every year due to politics proposing green energy productions as solutions to mitigate climate change effects. However, this kind of energy is ... [more ▼]

The number of wind turbines in the world grows significantly every year due to politics proposing green energy productions as solutions to mitigate climate change effects. However, this kind of energy is dependent on the weather. This implies that the wind production is irregular at a very short time scale. But the short time scale availability of the wind-based energy is important to the producers of energy as well as to the electric grid managers because the wind energy production can rise or fall rapidly which creates some financial and voltage variations. For these reasons, we study the past evolution of the availability of the wind quantity by analysing the intermittence of the wind speed in Belgium during the last 22 years (1989-2010). To reach this goal, we use the regional model WRF (Weather and Research Forecast model) developed by the UCAR community users. In a first time, the WRF model is forced by the NCEP2 reanalysis outputs to obtain a regionalisation of the weather conditions over a domain centred on Belgium at a resolution of 10 km. This resolution allows to capture the minute-based time scale variability of wind speed and consequently the irregular behaviour of the wind power production. In a second time, the WRF model is forced by the ERA-Interim reanalysis outputs with the same configuration. To obtain a value of the wind intermittence, we calculate the persistence of a wind blowing continuously with a minimum speed of 1 ms-1, then the persistence of a wind blowing continuously with a minimum speed of 2 ms-1, etc. The persistence of the wind speed and its evolution over 22 years are characterised by : (a) the mean wind speed over a fixed period (monthly, seasonally, … ), (b) the mean duration of a wind speed above x ms-1 over the same fixed period and (c) the evolution of (a) and (b) during the studied period. This analysis is made with the outputs of WRF-NCEP2 and with the outputs of WRF-Interim allowing to evaluate the impact of forcing fields into WRF-based wind climatology. [less ▲]

Detailed reference viewed: 103 (9 ULg)
Full Text
Peer Reviewed
See detailInfluences de l'environnement d'un parc éolien sur la prévision de sa production électrique à l'aide des modèles GFS (50km/3h) et WRF (2km/15min) : Le cas du parc éolien d'Amel (Haute-Belgique)
Doutreloup, Sébastien ULg; Fettweis, Xavier ULg; Erpicum, Michel ULg

in Fazzini, Massimiliano; Beltrando, Gérard (Eds.) Actes du XXIVème Colloque International AIC : Climat Montagnard et Risques (2011, September 06)

The economic and climate contexts require to use more electricity from wind farms. However this kind of production is intermittent, therefore it is necessary to forecast this resource at least 1 day ahead ... [more ▼]

The economic and climate contexts require to use more electricity from wind farms. However this kind of production is intermittent, therefore it is necessary to forecast this resource at least 1 day ahead. Our laboratory has developed a forecasting model of wind-based electricity generation based on a global meteorological model (GFS) with a resolution of 50 km and 3 h. But this model has a resolution too coarse for a wind farm. So we have configured the regional model WRF with resolution of 2 km and 15 min to obtain better forecasts. Finally, the WRF model provides better forecasts, but both must be adjusted to take into account the direct environment of the wind farm. [less ▲]

Detailed reference viewed: 65 (6 ULg)
Full Text
Peer Reviewed
See detailAnalysis of the past (1970-1999) and future (2046-2065 and 2081-2100) evolutions of precipitation and temperature, in the Province of Binh Thuan, South East Vietnam, based on IPCC models
Doutreloup, Sébastien ULg; Erpicum, Michel ULg; Fettweis, Xavier ULg et al

in Proceedings of the 1st International Conference on Energy, Environment and Climate Change (2011, August)

Natural resource-dependent societies in developing countries are facing increased pressures linked to global climate change. The Province of Binh Thuan, in South East Vietnam, where rainfall is on average ... [more ▼]

Natural resource-dependent societies in developing countries are facing increased pressures linked to global climate change. The Province of Binh Thuan, in South East Vietnam, where rainfall is on average 500 to 700 mm but can drop as low as 350 mm in some years, knows a recent increase of agricultural activities in order to contribute to reduce poverty although the technical efficiency of Binh Thuan is still very low. Within this framework of higher dependency of the local economy on the agricultural sector, there is growing evidence that changes in climate extremes are increasing exposure of currently vulnerable rural populations. In order to assess the future climate of the province of Binh Thuan, only three models able to simulate the current climate in the study area were used out of the 24 selected by the IPCC: CCCMA-T47, INGV and IPSL. The future climate projections (that is 2046-2065 and 2081-2100 compared to historical data 1970-1999) were focused on two targets: [i] assessing changes in climate statistics, and [ii] analysing the beginning and the end of the rainy season. [i] The first analysis indicates an increase of mean temperature of about 1.6°C (over 2046-2065) and 2.5°C (over 2081-2100) and an increase of extreme temperatures and extreme rainfall events. However, no significant changes about the evolution of the annual amount of precipitation were found. [ii] The second analysis indicates that the dry season is likely to be longer in 2046-2065 owing to a delay in the onset of the rainy season (up to 15 days) accompanied by an earlier end of the rainy season (up to 30 days). Although it must be kept in mind that precipitations are the most difficult climate variable to predict, it is likely that increasing water needs to support expending agriculture within the context of climate change in the Province of Binh Thuan will be a challenge. Indeed, extreme rainfall events are likely to increase and unchanged yearly amounts of precipitation should be concentrated in time. [less ▲]

Detailed reference viewed: 77 (14 ULg)
Full Text
See detailEvolution of wind behaviour and of its potential for wind power production in Belgium during the last 30 years.
Doutreloup, Sébastien ULg; Fettweis, Xavier ULg; Sacré, Bernard ULg et al

Poster (2011, April 04)

The number of wind turbines in the world grows significantly every year due to politics proposing green energy productions as solutions to mitigate climate change effects. However, this kind of energy is ... [more ▼]

The number of wind turbines in the world grows significantly every year due to politics proposing green energy productions as solutions to mitigate climate change effects. However, this kind of energy is dependent on the weather. This implies that the wind production is irregular at a very short time scale. But the short time scale availability of the wind-based energy is important to the producers of energy as well as to the electric grid managers because the wind energy production can rise or fall rapidly which creates some financial and voltage variations. For these reasons, we study the past evolution of the availability of the wind quantity by analysing the intermittence of the wind speed in Belgium during the last 30 years. To reach this goal, we use the regional model WRF (Weather and Research Forecast model) developed by the UCAR community users. The WRF model is forced by the NCEP2 Reanalysis model to obtain a regionalisation of the weather conditions over a domain centred on Belgium at a spatiotemporal resolution of 10 km and 1 min. This resolution allows to capture the minute-based time scale variability of wind speed and consequently the irregular behaviour of the wind power production. To obtain a value of the wind intermittence, we calculate the persistence of a wind blowing continuously with a minimum speed of 1 ms-1, then the persistence of a wind blowing continuously with a minimum speed of 2 ms-1, etc. The persistence of the wind speed and its evolution over 30 years are characterised by : (a) the mean wind speed over a fixed period (monthly, seasonally, … ), (b) the mean duration of a wind speed above x ms-1 over the same fixed period and (c) the evolution of (a) and (b) during the studied period. This study will show the evolution during the last decades of the wind behaviour in Belgium and its potential for electricity production. [less ▲]

Detailed reference viewed: 55 (8 ULg)
Full Text
See detailEvaluation over Greenland of WRF with GC-NET observations (1995-2005) by comparison with 2 other RCMs
Sacré, Bernard ULg; Fettweis, Xavier ULg; Doutreloup, Sébastien ULg et al

Conference (2011)

In the context of climate change, the Greenland Ice Sheet (GrIS) plays an important role in sea level variation and oceanic thermohaline circulation changes. Unfortunately, Global Climate Models do not ... [more ▼]

In the context of climate change, the Greenland Ice Sheet (GrIS) plays an important role in sea level variation and oceanic thermohaline circulation changes. Unfortunately, Global Climate Models do not illustrate enough the characteristics of Greenland. To solve that, specific RCMs have been developed to take into account the features of polar regions. In this project, we compare three RCMs : the MAR model, the RACMO model and the Weather Research and Forecasting (WRF) model. WRF is an open source model developed by the Mesoscale and Microscale Meteorology Division of NCAR. We use here the standard WRF (version 3.2.1) and its polar optimization (called polar WRF). The MAR version tuned for the GrIS and coupled with a 1D surface scheme called SISVAT (for Soil Ice Snow Vegetation Atmosphere Transfer) is compared here. The version of RACMO is a specific version for the Greenland climate, RACMO2/GR. This model contains a special surface module for snow-ice treatment and other modifications concerning, for example, the surface turbulence heat flux or the surface roughness. The comparison is made on a domain centered on Greenland at a 25-km horizontal resolution over the 1995-2005 period when Automatic Weather Station (AWS) measurements are available from the Greenland Climate NETwork (GC-NET). Statistics (mean, bias, RMSE, correlation coefficient) are calculated for the near-surface temperature, surface pressure, 10m-wind speed and specific humidity for winter (October to April) and summer (May to September). In addition, the modeled snowfall are evaluated with ice core-based snow accumulation climatologies. Comparison shows a significant improvement from RCMs compared to the reanalyses (NCEP2 and ERA-INTERIM) in respect to the AWS measurements. RACMO and MAR seem to compare better with observations than WRF. [less ▲]

Detailed reference viewed: 58 (7 ULg)
Full Text
See detailInterests of using a regional model to forecast wind power production
Doutreloup, Sébastien ULg

Poster (2010, November 05)

European policies have decided to reduce the greenhouse gas emissions of 20% and to reach 20% of renewable power production by 2020. Increasing wind power is one of the numerous solutions to reach these ... [more ▼]

European policies have decided to reduce the greenhouse gas emissions of 20% and to reach 20% of renewable power production by 2020. Increasing wind power is one of the numerous solutions to reach these goals. However, this kind of energy production depends on the meteorological conditions and gives it an intermittent behaviour. The wind speed variations cause voltage and frequency fluctuations that are unacceptable for the power grid. Therefore, forecasting production will become essential with the aim of integrating this kind of energy production into the power grid. We have developed and compared two forecasting models which give as outputs the wind power production every 15 minutes over the Belgian territory: the first one uses the outputs from the global model GFS (available at a horizontal resolution of 0.5° every 3h) and the second one uses the regional climate model WRF-NMM (using a horizontal resolution of 4km). Both of these models predict the wind speed and transform wind speed into wind power production, using a power curve which depends on the wind turbines and their characteristics. The first model using the GFS outputs is not precise enough in space and time to correctly forecast the wind speed in punctual wind farms. That is why we apply some specific tunings on these forecasts. These tunings depend on the air density, the wind direction and the stability of the air mass. The second model using the WRF-NMM outputs runs over the Belgian territory. Initial conditions are forced by the GFS outputs at 0.5° and WRF computes a physical based spatio-temporal downscaling of the meteorological variables. The outputs have a spatial resolution of 4 km and a time resolution of 15 minutes. Some tunings are also needed to adjust the wind power forecasts by comparison to the wind power observations. We present here some results of both models and the interest of using a regional model for more precise wind power forecasting. [less ▲]

Detailed reference viewed: 46 (6 ULg)
Full Text
Peer Reviewed
See detailInterests of regional modelisation for wind power forecasting
Doutreloup, Sébastien ULg; Fettweis, Xavier ULg; Erpicum, Michel ULg

Conference (2010, October 22)

European policies have decided to reduce the greenhouse gas emissions of 20% and to reach 20% of renewable power production by 2020. Increasing wind power is one of the numerous solutions to reach these ... [more ▼]

European policies have decided to reduce the greenhouse gas emissions of 20% and to reach 20% of renewable power production by 2020. Increasing wind power is one of the numerous solutions to reach these goals. However, this kind of energy production depends on the meteorological conditions and gives it an intermittent behaviour. The wind speed variations cause voltage and frequency fluctuations that are unacceptable for the power grid. Therefore, forecasting production will become essential with the aim of integrating this kind of energy production into the power grid. We have developed and compared two forecasting models which give as outputs the wind power production every 15 minutes over the Belgian territory: the first one uses the outputs from the global model GFS (available at a horizontal resolution of 0.5° every 3h) and the second one uses the regional climate model WRF-NMM (using a horizontal resolution of 4km). Both of these models predict the wind speed and transform wind speed into wind power production, using a power curve which depends on the wind turbines and their characteristics. The first model using the GFS outputs is not precise enough in space and time to correctly forecast the wind speed in punctual wind farms. That is why we apply some specific tunings on these forecasts. These tunings depend on the air density, the wind direction and the stability of the air mass. The second model using the WRF-NMM outputs runs over the Belgian territory. Initial conditions are forced by the GFS outputs at 0.5° and WRF computes a physical based spatio-temporal downscaling of the meteorological variables. The outputs have a spatial resolution of 4 km and a time resolution of 15 minutes. Some tunings are also needed to adjust the wind power forecasts by comparison to the wind power observations. We present here some results of both models and the interest of using a regional model for more precise wind power forecasting. [less ▲]

Detailed reference viewed: 58 (8 ULg)
Full Text
See detailAnalysis of the evolution of the climate parameters, especially precipitations and temperatures, in the province of Binh Thuan in Southern Vietnam based on IPCC models
Doutreloup, Sébastien ULg; Fettweis, Xavier ULg; Ozer, Pierre ULg

Report (2010)

This research is implied into the BELSPO / Vietnamese desertification project and the aim of this work is to analyse the future evolution of the temperatures and precipitations in the region of Binh Thuan ... [more ▼]

This research is implied into the BELSPO / Vietnamese desertification project and the aim of this work is to analyse the future evolution of the temperatures and precipitations in the region of Binh Thuan thanks to the IPCC models (CMIP3). [less ▲]

Detailed reference viewed: 45 (10 ULg)
See detailAmélioration d’un modèle de prévision quart horaire des productions des parcs éoliens d’Amel et de Perwez en Belgique.
Doutreloup, Sébastien ULg

Master's dissertation (2009)

Une société de distribution d'électricité a sollicité le Laboratoire de Climatologie (ULg) pour lui fournir, tous les jours à 8 heures du matin, la prévisions de la production électrique pour deux parcs ... [more ▼]

Une société de distribution d'électricité a sollicité le Laboratoire de Climatologie (ULg) pour lui fournir, tous les jours à 8 heures du matin, la prévisions de la production électrique pour deux parcs éoliens en Belgique, avec un pas de temps de 15 minutes. Depuis janvier 2008, un modèle opérationnel a été élaboré dans ce laboratoire pour calculer ces prévisions. Ce mémoire a contribué à l'amélioration de ce modèle de prévisions de production d'énergie électrique d'origine éolienne. Le modèle opérationnel mis en exploitation à la suite de ce mémoire reste relativement simple et ne fait intervenir que peu de ressources informatiques. En effet, le temps nécessaire pour calculer la prévision est très court car le modèle utilise directement les sorties du modèle global de prévisions météorologique « Global Forecast System » (GFS) sans downscaling, contrairement à ce qui est proposé dans cette thèse. Le but de ce mémoire était donc d'apporter des améliorations au modèle opérationnel et ainsi d'améliorer la qualité des prévisions de production rendues. Afin de créer ces améliorations et de les tester dans de bonnes conditions, nous avons, tout d'abord, trié nos données d'observations et de prévisions afin de nous affranchir des erreurs d'observations et de transmissions de la production des parcs éoliens et des imprécisions liées à l'utilisation de prévisions météorologiques. Les données d'observation sont parfois biaisées dues à l'arrêt des éoliennes, principalement pour cause de givre ou de maintenance. Les prévisions GFS utilisées dans le modèle opérationnel contenaient beaucoup d'erreurs liées à l'utilisation des prévisions météorologiques de la veille (00hTU) pour calculer les prévisions de production pour le lendemain. Pour calculer et mettre à l'épreuve les améliorations apportées au modèle, nous avons utilisé les prévisions de GFS les plus récentes possibles (toutes les 6 heures) de chaque prévision de production, et qui sont donc a priori les meilleures possibles. Une fois libéré de ces contraintes, nous avons alors testé une série d'éléments liés aux sites d'implantation des éoliennes afin d'améliorer les prévisions des productions des deux parcs. Différents éléments d'amélioration ont été retenus : - Une rose des rugosités : Cette amélioration a permis de corriger la vitesse du vent modélisée selon la direction d'où il provenait, suivant la présence d'obstacles de natures diverses (des arbres, une colline, une vallée, ...) - Une modification de la production en fonction de la densité de l'air qui influence la productivité des éoliennes pour une même vitesse de vent. - Un calibrage par rapport à l'indice K-Index qui caractérise l'instabilité de l'atmosphère et donc les rafales de vent dues aux mouvements convectifs locaux qui ne sont pas modélisés par GFS. - Un calibrage par rapport aux amplitudes thermiques diurnes qui influencent les vents locaux non modélisés par GFS. Au final, nous avons pu améliorer les prévisions de l'ordre de 20% pour le premier parc et de l'ordre de 5% pour le second par rapport au modèle de base. La faible amélioration pour le second parc provient du fait que les prévisions étaient déjà de bonne qualité dans le modèle opérationnel originel et que les calibrations empiriques contenues dans ce modèle suffisaient, contrairement contrairement au premier parc dont la physionomie du terrain à proximité des éoliennes était très complexe. [less ▲]

Detailed reference viewed: 113 (12 ULg)