Elsevier

Atmospheric Research

Volume 247, 1 January 2021, 105157
Atmospheric Research

Sensitivity analysis of the WRF model: Assessment of performance in high resolution simulations in complex terrain in the Canary Islands

https://doi.org/10.1016/j.atmosres.2020.105157Get rights and content

Highlights

  • T2m showed great correlation between observation and forecast.

  • A systematic error was detected for T2m on the eastern coasts of the islands.

  • The PBL scheme chosen had a great impact in precipitation.

  • Wind gust showed a systematic negative bias.

  • The model presented inability to forecast wind gust higher 100 km/h.

Abstract

Canary Islands and other regions have been greatly damaged by weather events during the last decades. For this reason, the main duty of National Meteorological Services is to minimize socio-economic losses by forecasting adverse weather episodes with enough time in advance. To achieve this goal, the use of numerical weather prediction models is highly relevant. And, even more crucial, is to comprehend the model accuracy.

In this paper, an exhaustive sensitivity analysis of the Weather Research and Forecasting (WRF) model over the Canary Islands has been carried out. The complex terrain of the archipelago makes the islands a test bench of high interest. Four scores were used to assess the accuracy of the model configurations: Bias, mean absolute error (MAE), root of the Mean Squared Error (RMSE), and the correlation coefficient (r). Initially, twenty-five WRF model configurations were considered. However, a preliminary test discarded inadequate configurations, and reduced the number to six. The variables of interest were air temperature at 2 m (T2m), maximum 1-h wind gust at 10 m and 3-h rainfall accumulation. The results indicated a systematic wind speed underestimation. This underestimation is related to the influence of the location and the complex orography. The most accurate wind forecasts were obtained using the Mellor-Yamada-Janjic Planetary Boudary Layer (PBL) scheme with the WSM6 microphysics (MP) scheme. Another major conclusion is that, for precipitation, the PBL scheme has a greater impact than the MP scheme. Finally, the results show that the Boulac – Thompson combination is the most accurate regarding T2m forecast.

Introduction

Extreme weather events have negative impacts on transportations and communications, consequently resulting in catastrophic effects on distinct aspects of people's lives and economy. Despite the apparent climatic mildness, the frequency and intensity of the severe weather events have serious consequences on the Canary Islands (Dorta, 2007). To understand this impact, statistics show that severe weather events caused 74 fatalities between 1995 and Eiserloh, 2014 (Suárez-Molina et al., 2018). According to the CCS (“Consorcio de Compensación de Seguros”, a public organization funded by the Ministry of Economy, Industry and Competitiveness of the Spanish Government), between 1996 and 2018, floods and windstorms in the Canary Islands produced more than 211 million Euros in losses (Suárez-Molina et al., 2020).

The accuracy of Numerical Weather Prediction (NWP) models in complex terrain is lower than over flat and homogeneous terrain. This discrepancy is attributed to the fact that boundary-layer processes in complex terrain are not well represented by NWP models. Previous studies have evaluated the performance of different planetary boundary layer (PBL) parameterization schemes in locations known for complex atmospheric situations (Pérez et al., 2006; Bossioli et al., 2009). Microphysics schemes in numerical models play a key role in simulating the formation of cloud droplets, precipitation, and land surface temperature. It also takes into account the interactions and energy fluxes between the atmosphere and the surface, which is considered a key parameter in many hydrological, meteorological and environmental studies (Anderson et al., 2011).

In recent years, the use of the Weather Research and Forecasting (WRF) model in operational mode has increased. For instance, since 2017, the National Centers for Environmental Prediction (NCEP) use in operational mode the Hurricane Weather Research and Forecast (HWRF) system (Biswas et al., 2018). In addition, the WRF model has been used by other authors in operational mode with different purposes (Hsiao et al., 2012; Hamill, 2014; Sahoo et al., 2019). The WRF model is also being used in γSREPS, an Ensemble Prediction System developed by AEMET, the Spanish Meteorological Agency (Callado et al., 2019).

Although other studies have used WRF in the Canary Islands for particular phenomena (Marrero et al., 2008; Jorba et al., 2015; Quitián-Hernández et al., 2018), a comprehensive sensitivity analysis has not been carried out before. Such study is essential to determine the most convenient model setup for this geographical domain (Borge et al., 2008). In addition, it should be taken into consideration that the operational forecast of convective episodes is more problematic in subtropical regions such as the Canary Islands (Žagar et al., 2005).

The purpose of this work is to evaluate the quality of WRF forecasts in the Canary Islands. The fields analyzed–air temperature at 2 m (T2m), maximum 1-h wind gust at 10 m and 3-h rainfall accumulation–are of vital importance for issuing meteorological warnings (METEOALERTA, 2018). The period analyzed (15 days in February 2018) includes various weather patterns; therefore, the sensitivity analysis will evaluate the model performance under different atmospheric conditions.

This paper is structured as follows: Section 2 describes the study area, the configuration of the WRF experiments and the observational dataset used to evaluate the model performance. Section 3 presents the results of the performance evaluation. Finally, Section 4 summarizes the paper conclusions.

Section snippets

Study area and dataset

This research is focused on the Canary Islands (Fig. 1). This archipelago is in front of the west coast of North Africa in the subtropical zone (27°37′–29°25′N and 18°10′–13°20′W). The archipelago is formed by seven islands of volcanic origin that present a complex orography. The highest point is Mount Teide (3718 m) on Tenerife (TF). With Tenerife being by far the highest island, La Palma (LA), Gran Canaria (GC), La Gomera (GO) and El Hierro (HI) constitute a medium cluster with highest

Results and discussion

The results section is structured as follows; in Section 3.1 we will present the results of the sensitivity analysis where we discuss the different scores for the six configurations; then, we will study the dependency to the lead time (Section 3.2), to the location (Section 3.3), and altitude (Section 3.4).

Conclusions

Sensitivity analysis can be used to determine optimum WRF model configurations. However, the best configuration depends on location and meteorological conditions. In this research, a comprehensive sensitivity analysis has been carried out in the Canary Islands. Due to the impact on the socio-economic activities, the analysis has been focused on T2m, wind gust and 3-h rainfall accumulation. Subsequently, the most relevant conclusions are summarized:

  • This research allowed discarding inappropriate

Author statement

The paper entitled “Sensitivity analysis of the WRF Model: Assessment of performance in high resolution simulations in complex terrain in the Canary Islands” was carried out by the authors David Suárez-Molina, Sergio Fernández-González, Gustavo Montero, Albert Oliver and Juan Carlos Suárez González. In the following lines, the contribution of each author will be detailed:

David Suárez-Molina: He is the main contributor to this paper, being the responsible of the conceptualization, design of the

Declaration of Competing Interest

The authors claim that there is no conflict of interest, nor any funding source that intercedes with the free publication of results obtained in this research.

Acknowledgements

Observational data are provided by the State Meteorological Agency of Spain (AEMET). The authors are grateful to the Weather Forecast Research Team for developing “Verif” software. Special thanks go to the SAFEFLIGHT (CGL2016-78702-C2-1-R and CGL2016-78702-C2-2-R) and UE ERA-NET Plus NEWA (PCIN2016-080) projects.

References (61)

  • M.K. Biswas et al.

    Hurricane Weather Research and Forecasting (HWRF) Model: 2017 Scientific Documentation (No. NCAR/TN-544+STR)

    (2018)
  • E. Bossioli et al.

    The role of planetary boundary-layer parameterizations in the air quality of an urban area with complex topography

    Bound. Layer Meteorol.

    (2009)
  • P. Bougeault et al.

    Parameterization of orography-induced turbulence in a mesobeta--scale model

    Mon. Weather Rev.

    (1989)
  • Callado, A., Escribà, P., Gil, D., Compte., M, Martínez, M., González, S., Quintero, D., García-Moya, J.A., 2019....
  • L. Cana et al.

    A Numerical Study of a Sea Breeze at Fuerteventura Island, Canary Islands, Spain

    Bound.-Layer Meteorol.

    (2020)
  • J. Carrillo et al.

    Characterization of the marine boundary layer and the trade-wind inversion over the sub-tropical North Atlantic

    Boundary-Layer Meteorol.

    (2016)
  • F. Chen et al.

    Coupling an advanced land surface–hydrology model with the Penn State–NCAR MM5 modeling system. Part I: model implementation and sensitivity

    Mon. Weather Rev.

    (2001)
  • W.Y.Y. Cheng et al.

    Evaluation of Surface Sensible Weather forecasts by the WRF and the Eta Models over the Western United States

    Weather Forecasting

    (2005)
  • F.K. Chow et al.

    Mountain Weather Research and Forecasting

    (2013)
  • B.A. Colle et al.

    The 13–14 December 2001 IMPROVE-2 event. Part III: simulated microphysical budgets and sensitivity studies

    J. Atmos. Sci.

    (2005)
  • A. Dandou et al.

    Investigation of turbulence parametrization schemes with reference to the atmospheric boundary layer over the Aegean Sea during etesian winds

    Boundary-Layer Meteorol.

    (2017)
  • P. Dorta

    Catálogo de riesgos climáticos en Canarias: Amenazas y vulnerabilidad

    Geographicalia

    (2007)
  • J.P. Evans et al.

    Evaluating the performance of a WRF physics ensemble over South-East Australia

    Clim. Dynam.

    (2012)
  • S. Fernández-González et al.

    Numerical simulations of snowfall events: Sensitivity analysis of physical parameterizations

    J. Geophys. Res.-Atmos.

    (2015)
  • S. Fernández-González et al.

    Uncertainty quantification and predictability of wind speed over the Iberian Peninsula

    J. Geophys. Res.-Atmos.

    (2017)
  • S. Fernández-González et al.

    Sensitivity analysis of the WRF model: wind-resource assessment for complex terrain

    J. Appl. Meteorol. Clim.

    (2018)
  • M.F. Garvert et al.

    The 13–14 December 2001 IMPROVE-2 event. Part II: Comparisons of MM5 model simulations of clouds and precipitation with observations

    J. Atmos. Sci.

    (2005)
  • S. González Rocha et al.

    Planet Boundary Layer Parameterization in Weather Research and forecasting (WRFv3.5): Assessment of Performance in High Spatial Resolution Simulations in complex Topography of Mexico

    Computacion y Sistemas.

    (2017)
  • T.M. Hamill

    Performance of Operational Model Precipitation Forecast Guidance during the 2013 Colorado Front-Range Floods

    Mon. Weather Rev.

    (2014)
  • C. Hohenegger et al.

    Predictability mysteries in cloud-resolving models

    Mon. Weather. Rev.

    (2006)
  • Cited by (0)

    View full text