The Mesoscale Compressible Community (MC2) model [1], devoted for weather forecasting and used in the Wind Energy Simulation Toolkit (WEST) [2], performs well for simulations over flat, gentle and moderate terrain slopes but is subject to numerical instability and strong spurious flows in presence of steep topography. To remove its inherent computational mode and reduce the wind overestimation due to terrain-induced numerical noise, a new semi-implicit (N-SI) scheme [3] was implemented to discretize and linearize the non-hydrostatic Euler equations with respect the mean values of pressure and temperature instead of arbitrary reference state values, redefining as well the buoyancy to use it as the thermodynamic prognostic variable. Additionally, the climate-state classification of the statistical-dynamical downscaling (SDD) method [4] is upgraded by including the Brunt-Väisälä frequency that accounts for the atmospheric thermal stratification effect on wind flow over topography. The present study provides a real orographic flow validation of these numerical enhancements in MC2, assessing their individual and combined contribution for an improved initialization and calculation of the surface wind in presence of high-impact terrain. By statistically comparing the wind simulations with met-mast data, obtained within the Whitehorse area of the Canadian Rocky Mountains, it is confirmed that these numerical enhancements may reduce over 40 percent of the wind overestimation, thus, attaining more accurate results that ensure reliable wind resource assessments over complex terrain.

Due to the natural variability of the wind, it is necessary to conduct thorough wind resource assessments to determine how much energy can be extracted at a given site. Lately, important advancements have been achieved in numerical methods of multiscale models used for high resolution wind simulations over steep topography. As a contribution to this effort, an enhanced numerical method was devised in the mesoscale compressible community (MC2) model of the Meteorological Service of Canada, adapting a new semi-implicit scheme with its imbedded large-eddy simulation (LES) capability for mountainous terrain. This implementation has been verified by simulating the neutrally stratified atmospheric boundary layer (ABL) over flat terrain and a Gaussian ridge. These preliminary results indicate that the enhanced MC2-LES model reproduces efficiently the results reported by other researchers who use similar models with more sophisticated sub-grid scale turbulence schemes. The proposed multiscale method also provides a new wind initialization scheme and additional utilities to improve numerical accuracy and stability. The resulting model can be used to assess the wind resource at meso- and micro-scales, reducing significantly the wind speed overestimation in mountainous areas.

In-cloud icing can impose safety concerns and economic challenges for various industries. Icing climate representations proved beneficial for optimal designs and careful planning. The current study investigates in-cloud icing, its related cloud microphysics and introduces a 15-year time period climatology of icing events. The model was initially driven by reanalysis data from North American Regional Reanalysis and downscaled through a two-level nesting of 10 km and 5 km, using a limited-area version of the Global Environment Multiscale Model of the Canadian Meteorological Center. In addition, a hybrid approach is used to reduce time consuming calculations. The simulation realized exclusively on significant icing days, was combined with non-significant icing days as represented by data from NARR. A proof of concept is presented here for a 1000 km area around Gaspe during January for those 15 years.An increase in the number and intensity of icing events has been identified during the last 15 years. From GEM-LAM simulations and within the atmospheric layer between 10 m and 200 m AGL, supercooled liquid water contents indicated a maximum of 0.4 g m(-3), and 50% of the values are less than 0.05 g m(-3). All values of median volume diameters (MVD) are approximately capped by 70 pm and the typical values are around 15 pm. Supercooled Large Droplets represent approximately 5%. The vertical profile of icing climatology demonstrates a steady duration of icing events until the level of 60 m. The altitudes of 60 m and 100 m indicate substantial icing intensification toward higher elevations. GEM-LAM demonstrated a substantial improvement in the calculation of in-cloud icing, reducing significantly the challenge posed by complex terrains. (C) 2014 Elsevier B.V. All rights reserved.

The present study aims at producing localized near-surface icing envelopes that contribute to optimal designs of near-surface structures under icing conditions. For this purpose, the 99th percentiles of liquid water content and wind speed that correspond to low-level supercooled clouds are used. Near-surface icing events and cloud microphysics are explored using North American Regional Reanalysis during winter months (D–J–F), over 32years. The investigation of the regional climatology of icing events involves 74 regional zones that cover Canada. For each zone, regional power loss of wind turbines under icing condition is estimated. The East and West coastal regions of the Hudson Bay demonstrate higher liquid cloud water of 0.4gm−3. The highest potential of wind is located in the West coast. The climatology of liquid water content around the Rocky Mountains manifests orographic condensation of the Pacific moisture transported toward the mountains. Within the range of temperatures [−15°C to 0°C], the near-surface results over Canada show that the monthly mean of wind speeds varies mostly between 4ms−1 and 10ms−1, and the mean supercooled cloud water content decreases linearly from 0.3gm−3 to 0.2gm−3, with decreasing temperature. The quantification of ice accumulation and the duration of icing events reveal that the West and the South of the Hudson Bay as well as the North of Manitoba and Ontario are exposed to extreme icing conditions, with a monthly accumulation that varies from 150mm to 225mm, and a monthly duration of icing events near 375h. Over the region encompassing the Gaspe Peninsula, St. Lawrence River and New Brunswick the higher limit of wind speed varies around 15ms−1. The cloud water upper limit of 0.45gm−3 occurs in December and January, and 0.3gm−3 in February. With decreasing temperature these upper limits reach 0.1gm−3 at −15°C. The highest wind energy is located in the Canadian East Coast regions. The Canadian arctic is characterized mostly by lower wind energy and larger power degradation under icing conditions. The average of wind turbine power loss during winter months (D–J–F) under icing conditions varies throughout Canada and reaches its maximum of about 15% over North-East of Manitoba, South of the Hudson Bay and the North coast of Ontario.

Environment Canada (EC) and Hydro-Québec (HQ) have been collaborating in a Research & Development and Demonstration project on a high resolution wind energy dedicated forecasting system (SPÉO: Système de Prévision ÉOlien under its French acronym). This project emphasizes the operational tests and the forecast of high impact events, e.g. wind ramps. It was found that SPÉO improves the Canadian Regional Deterministic Prediction System (RDPS), by about 18% in terms of the RMSE (Root Mean Square Error) of the predicted wind speed when compared with mast observations from three wind power plants. The improvement is most significant in the cold season. When the average wind speed measured at all wind turbines (nacelle anemometer) is used as a reference, SPÉO improves the RMSE of the average wind speed at a wind power plant in complex terrain (24%) compared with that of RDPS. However, there is almost no improvement for two other wind power plants located in less complex terrain. The average wind speed is corrected with the average wind speed measured at all turbines, and is then fed into a wind-to-power conversion module for power production forecasts. The power production forecast is improved by 6% on average in complex terrain when SPÉO winds are used as input compared to the RDPS. The most important finding of this project is SPÉO's ability to predict ramps due to mountain waves/downslope winds. The proposed forecast index for ramps based on the Froude number is useful for predicting the onset of this kind of ramp when a high resolution NWP model is unavailable.

Wind turbine performance depends mainly on the wind speed and aerodynamics of blades. The roughness generated from ice accretion can significantly reduce the aerodynamics and consequently the power production of the wind turbine. This study locates the glaze and rime ice on the blade, to detect the critical zones involved in significant power production loss. On the blade, the distribution of the elementary power production as well as the type and thickness of the accreted ice are inconsistent. Under icing conditions, the outer section of the blade starting from the radial position r/R=0.8 contribute significantly to the blades aerodynamics. The freezing fraction is unevenly distributed; since it initially forms rime ice near the root and then glaze toward the tip of the blade. The critical freezing fraction 0.88 associated with the double horn ice shape is spatially limited and occupies a restricted segment on the blade and gradually moves towards the tip with decreasing temperature. With the use of power degradation analogy with sub-scaled rotor blades of a helicopter under icing conditions, a power loss factor is introduced to quantify and locate power loss along the blades of wind turbines. The study is based on four values of liquid water content that delineate five classes of icing severity. Including power loss factor, the most significant power loss that corresponds to freezing fraction 0.88 is found to be located at r/R~[0.93 0.96] which corresponds to T=−2.6°C, −4.5°C, −12°C, and −20°C and for liquid water content LWC=0.04g·m−3, 0.07g·m−3, 0.2g·m−3, and 0.36g·m−3 respectively. The resulted power degradation can reach a maximum of 40%. Locally it is the shape rather than the thickness of ice that causes more power loss, meanwhile when considering the whole blade, power degradation is controlled mainly by ice thickness regardless of the type of ice. The results obtained can help the setup of a sensor that triggers the ice-protection system upon detection of critical freezing fraction.

The large size of modern wind turbines and wind farms triggers processes above the surface layer, which extend to the junction between microscales and mesoscales, and pushes the limits of existing approaches to predict the wind. The main objectives of this study are thus to introduce and evaluate an approach that will better account for physical processes within the atmospheric boundary layer (ABL), and allow for both microscale and mesoscale modeling. The proposed method, in which mathematical model and main numerical aspects are presented, combines a mesoscale approach with a large-eddy simulation (LES) model based on the Compressible Community Mesoscale Model (MC2). It is evaluated relying on a shear-driven ABL case allowing the authors to assess the model behavior at very high resolution as well as more specific numerical aspects such as the vertical discretization and time and space splitting of turbulence-related terms. The proposed LES-capable mesoscale model is shown to perform on par with other similar reference LES models, while being slightly more dissipative. A new vertical discretization of the turbulent processes eliminates a spurious numerical mode in the solution. Finally, the splitting of horizontal and vertical turbulence-related terms is shown to have no impact on the results of the test cases. It is thus demonstrated that the revised MC2 is suitable at both microscales and mesoscales, thus setting a strong foundation for future work.

Atmospheric icing became a primary concern due to the significant impact and hazardous conditions of its accretion on structures. The objective of this study is to provide a map of icing events over 32 years (1979 to 2010) that describes the severity of winter icing. This information will prove useful to prevent damages and economical losses due to icing events by documenting the risk factor.To validate the icing climatology method, two case studies involving two topographically contrasting sites were selected: a simple terrain site which is the airport of Bagotville, near Saguenay (Canada) and a complex terrain site located in Mt Belair, near Quebec City (Canada). Ice accumulation calculated by the use of reanalysis data was quantified using ice accretion on a cylinder model. Comparison between measurement and the model over Bagotville revealed insignificant differences in ice accumulation less than 03 mm, and in duration of icing events less than 02 day. On the other hand, during winter months, the calculation that showed a maximum of 60 mm in January 1999 over Mt Belair site also had an underestimation of ice accumulation that varies from 5 mm to 16 mm. The horizontal resolution of NARR imposes a challenge on the calculation of icing events over complex terrains, especially during the months of November and March when air temperature is near freezing point. Taking into account the liquid water content, the duration of icing events and the classes of icing events as weighting factors, the icing severity index based on reanalysis data was introduced to assess the severity level of icing events, covering the north-east of Quebec including Quebec City, Sept-Iles, the east of Saguenay, the lower St Lawrence River and the Gaspe region. Consequently, an icing severity index mapping that represents the climatology of in-cloud atmospheric icing was produced. (c) 2013 Elsevier B.V. All rights reserved.

The authors evaluate the performance of current regional models in an intercomparison project for a case of explosive secondary marine cyclogenesis occurring during the Canadian Atlantic Storms Project and the Genesis of Atlantic Lows Experiment of 1986. Several systematic errors are found that have been identified in the refereed literature id prior years. There is a high (low) sea level pressure bias and a cold (warm) tropospheric temperature error in the oceanic (continental) regions. Though individual model participants produce central pressures of the secondary cyclone close to the observed during the final stages of its life cycle, systematically weak systems are simulated during the critical early stages of the cyclogenesis; Additionally, the simulations produce an excessively weak (strong) continental anticyclone (cyclone); implications of these errors are discussed in terms of the secondary cyclogenesis. Little relationship between strong performance in predicting the mass held and skill in predicting a measurable amount of precipitation is found. The bias scores in the precipitation study indicate a tendency for all models to overforecast precipitation. Results for the measurable threshold (0.2 mm) indicate the largest gain in precipitation scores results from increasing the horizontal resolution from 100 to 50 km, with a negligible benefit occurring as a consequence of increasing the resolution from 50 to 25 km. The importance of a horizontal resolution increase from 100 to 50 km is also generally shown for the errors in the mass field. However, little improvement in the prediction of the cyclogenesis is found by increasing the horizontal resolution from 50 to 25 km.

Although the Mesoscale Community Compressible (MC2) model successfully reproduces the wind climate (for wind energy development purposes) of the Gaspe region, equivalent simulations for the steep mountainous southern Yukon have been unsatisfactory. An important part of the problem lies in the provision of suitable boundary conditions in the lower troposphere. This paper will describe an alternative provision of boundary conditions to the MC2 model based partly on standard National Centers for Environmental Prediction/National Center for Atmospheric Research (NCEP/NCAR) Reanalysis statistics, however, with modified lower tropospheric conditions based on local radiosonde measurements.The MC2 model is part of the AnemoScope wind energy simulation toolkit which applies statistical-dynamical downscaling of basic large-scale weather situations (i.e., the NCEP/NCAR Reanalysis) to simulate the steady-state wind climate of a complex region. A case study summarized here imposes a typical mean winter temperature inversion on the boundary conditions to reduce downward momentum transfer in the MC2 model over the Whitehorse region. In conjunction with this step, the geostrophic wind at the boundaries is held constant (with height) in speed and direction, based on the (observed) dominant southwesterly winds above the mountaintops. The resulting simulation produces wind directions within the modelled domain that are in much better agreement with the available measurements. However, despite the imposed atmospheric stability, downward transfer of horizontal momentum from aloft still appears to exceed that occurring in nature.It is recommended that (in future studies of this type regarding mountain wind climate) the input statistics processed from the NCEP/NCAR Reanalysis be modified by referencing the geostrophic winds to a level above the mountaintops. It is also suggested that converting to a height (z) coordinate system may reduce the erroneous downward momentum transfer found in the present terrain-following grid.

Wind assessment in a coastal environment remains a complex issue for both synthetic aperture radar (SAR) satellite imagery and numerical weather prediction (NWP) models. This study compares the accuracy of each technique to improve the overall mapping precision of both methods. On the one hand, 14 RADARSAT-1 scenes of the St. Lawrence River (Canada) are converted into wind speeds using a hybrid model function that consists of the CMOD-IFR2 geophysical model function and a C-band polarization ratio. A priori information on wind directions is gathered from QuikSCAT scatterometer and in situ wind data. On the other hand, co-located wind maps are generated with the Environment Canada mesoscale compressible community (MC2) model. Comparisons between these two methods are then presented according to three approaches: a systematic SAR and MC2 comparison at 4 km grid-point spacing, a validation with observations (buoy and QuikSCAT scatterometer), and a local analysis of SAR and MC2 winds along a transect perpendicular to the coastline. The main features of the offshore wind fields are well resolved by both methods. The comparison study shows that SAR and MC2 winds have good agreement, with a root mean square difference for wind speeds of 2.07 m/s and a bias of 0.13 m/s.

A state-of-art wind mapping system, the Wind Energy Simulation Toolkit (WEST), was developed in the Meteorological Service of Canada (MSC) for use by the wind energy industry. WEST is based on a statistical-dynamical downscaling approach, i.e. (i) a statistical analysis of climate data to determine the basic atmospheric states, and (ii) a dynamic adaptation of each basic state to high-resolution terrain and surface roughness by using mesoscale and microscale models. The approach has already been used by Frank and Landberg (1997), in their KAMM/WAsP method, to create a numerical wind atlas. The novel part of WEST is the fixed wind-speed interval in classification scheme and the integration of different modules (meso-/micro-scale models and statistical module) into a single toolkit in a more portable form. WEST was built for use by industries not having sophisticated computer facilities. WEST is applied to the Gaspé region of Canada. The mesoscale model MC2 (operated within WEST) is run at 5 km resolution, while the microscale model within WEST is at 200 m resolution. The simulation results are evaluated in comparison with tower observations at a height of 40 m above ground level. The mean of the 29 observed sets of wind data is 6.6 m/s. The mean absolute difference between the observed and simulated sets of wind data is 0.83 m/s with MC2 (meso-component of WEST) and 0.69 with ‘full WEST’ (with both meso- and micro- components). The correlation coefficient of the mean wind-speeds between the simulations and observations for the 29 stations is improved from 0.5 with MC2 to 0.7 with WEST.

The improvement of Quantitative Precipitation Forecast (QPF) in mountainous area was the central supporting objective of the MAP project P1 devoted to the study of orographic precipitation. This paper attempts to review the main MAP-related achievements towards QPF improvement and to highlight the MAP-impact for developing QPF research and operational strategies.

The wind climate of the mountainous terrain in the southern Yukon is simulated using the Wind Energy Simulation Toolkit (WEST) developed by the Recherche en Prevision Numerique (RPN) group of Environment Canada and is compared to measurements in the field. WEST combines two models that operate at different spatial scales. The Mesoscale Compressible Community (MC2) model is a mesoscale numerical weather prediction model that produces simulations over large domains of the order of a thousand kilometres. The MC2 model uses long-term synoptic scale wind climate data from the analysis of radiosonde and other observations to simulate mean wind fields at tens of metres above the ground using a horizontal resolution of a few kilometres. The mesoscale results are used as input to MS-Micro/3 (Mason and Sykes (1979) version of the Jackson and Hunt (1975) model version for microcomputers/3-dimensional; MS-Micro hereafter), a more computer-efficient, microscale model with simpler linearized momentum equations and a domain restricted to a few tens of kilometres with horizontal grid sizes of tens or hundreds of metres. MS-Micro provides wind field results at specific wind generator hub heights (typically 30 to 50 in above ground level (AGL)) which are of interest to researchers and developers of wind farms.WEST shows relatively strong correlations between its simulated long-term mean wind speed and the measurements from ten wind energy monitoring stations. However, in the mountainous terrain of the Yukon, WEST tends to predict wind speeds which art about 40% too high. The model also produces erroneous wind directions and some were perpendicular to valley orientations. The most likely cause of the wind speed and direction errors is the substantially modified 5-km grid-spaced mesoscale terrain used in MC2. The WEST simulation was also found to double the wind speeds observed at airport stations and there was poor correlation between the simulated and observed wind speeds.The bias in the model could be attributed to a number of factors, including the use of smoothed topography by the model, the discrepancy between the neutral atmosphere assumed in MS-Micro and the normally observed stable atmosphere, the application of MS-Micro to every third grid point of the MC2 output, abnormally high sea level wind speeds in the input climate data for MC2, and a certain degree of disagreement between the land surface characteristics used in the model and those found in the field.At comparatively low computer cost, WEST predicts a wind climate map that compares favourably to the wind measurements made in several locations in the Yukon. However, the problem of the modified terrain in the mountainous regions is the most pressing problem and needs to be addressed before WEST is used in the mountainous regions of Canada.

During the Special Observation Period (SOP, 7 September–15 November, 1999) of the Mesoscale Alpine Programme (MAP), the Canadian Mesoscale Compressible Community Model (MC2) was run in real time at a horizontal resolution of 3 km on a computational domain of 350☓300☓50 grid points, covering the whole of the Alpine region. The WATFLOOD model was passively coupled to the MC2; the former is an integrated set of computer programs to forecast flood flows, using all available data, for catchments with response times ranging from one hour to several weeks. The unique aspect of this contribution is the operational application of numerical weather prediction data to forecast flows over a very large, multinational domain. An overview of the system performance from the hydrometeorological aspect is presented, mostly for the real-time results, but also from subsequent analyses. A streamflow validation of the precipitation is included for large basins covering upper parts of the Rhine and the Rhone, and parts of the Po and of the Danube. In general, the MC2/WATFLOOD model underestimated the total runoff because of the under-prediction of precipitation by MC2 during the MAP SOP. After the field experiment, a coding error in the cloud microphysics scheme of MC2 explains this underestimation to a large extent. A sensitivity study revealed that the simulated flows reproduce the major features of the observed flow record for most of the flow stations. The experiment was considered successful because two out of three possible flood events in the Swiss-Italian border region were predicted correctly by data from the numerical weather models linked to the hydrological model and no flow events were missed. This study has demonstrated that a flow forecast from a coupled atmospheric-hydrological model can serve as a useful first alert and quantitative forecast. Keywords: mesoscale atmospheric model, hydrological model, flood forecasting, Alps

This paper describes the set-up and application of a non-hydrostatic Canadian meteorological numerical model (MC2) for mesoscale simulations of wind field and other meteorological parameters over the complex terrain of Hong Kong. Results of the simulations of one case are presented and compared with the results of radiosonde and aircraft measurements. The model is proved capable of predicting high-resolution, three-dimensional fields of wind and other meteorological parameters within the Hong Kong territory, using reasonable computer time and memory resources.

The Canadian mesoscale model (Benoit et al., 1997) provided daily forecasts across the Alps at 3-km resolution during the entire MAP field phase of 1999. The model had been extensively optimized in previous years for efficiency on various computer architectures (Thomas et al., 1997) and accuracy inside as well as at the model boundaries (Thomas et al. 1998). An overview of its performance during MAP is given in Benoit et al. (2002). Following the experiment, it became more and more evident that there was a problem related to finescale orography forcing in the model (Schaer et al., 2002). The problem is soon to be explained (Klemp et al., 2002) and no doubt corrected. Here we describe a modification to the model dynamics kernel which in particular greatly reduces its spurious sensitivity to finescale orography.

The roles and mechanisms of transport and chemical transformation affecting surface ozone in the Canadian southern Atlantic region (SAR) are investigated using a three‐dimensional Eulerian comprehensive modeling system. The investigation is focused a regional ozone episode over the eastern North America during the first week of August 1988. The model performance is evaluated with available observations during this period. The model is shown to reproduce the general features of this regional episode, with better performance for the sites with clear local photochemical activities than for those strongly influenced by complex coastal flow. It is shown from the reconstructed time history of various processes along the backward trajectories originating from a site in Nova Scotia that the elevated ground‐level ozone in the SAR during the study period was associated with transport at low levels, under a strong southwesterly flow. The high ozone brought to the site was mostly produced within the stable marine boundary layer from precursors picked up earlier over the emission area on the east coast of the United States. A second transport situation was also simulated and shown to be associated with transport aloft that was never mixed to the surface. This study also shows that differential advection, due to stable stratification and vertical shear of horizontal wind, can deform surface‐based plumes to produce elevated layers of pollutants over the Gulf of Maine.