Search

The Arctic is a vital component of the global climate, and its rapid environmental evolution is an important element of climate change around the world. To detect and diagnose the changes occurring to the coupled Arctic climate system, a state-of-the-art synthesis for assessment and monitoring is imperative. This paper presents the Arctic System Reanalysis, version 2 (ASRv2), a multiagency, university-led retrospective analysis (reanalysis) of the greater Arctic region using blends of the polar-optimized version of the Weather Research and Forecasting (Polar WRF) Model and WRF three-dimensional variational data assimilated observations for a comprehensive integration of the regional climate of the Arctic for 2000–12. New features in ASRv2 compared to version 1 (ASRv1) include 1) higher-resolution depiction in space (15-km horizontal resolution), 2) updated model physics including subgrid-scale cloud fraction interaction with radiation, and 3) a dual outer-loop routine for more accurate data assimilation. ASRv2 surface and pressure-level products are available at 3-hourly and monthly mean time scales at the National Center for Atmospheric Research (NCAR). Analysis of ASRv2 reveals superior reproduction of near-surface and tropospheric variables. Broadscale analysis of forecast precipitation and site-specific comparisons of downward radiative fluxes demonstrate significant improvement over ASRv1. The high-resolution topography and land surface, including weekly updated vegetation and realistic sea ice fraction, sea ice thickness, and snow-cover depth on sea ice, resolve finescale processes such as topographically forced winds. Thus, ASRv2 permits a reconstruction of the rapid change in the Arctic since the beginning of the twenty-first century–complementing global reanalyses. ASRv2 products will be useful for environmental models, verification of regional processes, or siting of future observation networks.

Total 550 nm aerosol optical depth, surface fine particulate matter (PM2.5), and meteorological observations were assimilated with continuously cycling three-dimensional variational (3DVAR), ensemble square root Kalman filter (EnSRF), and hybrid variational-ensemble data assimilation systems. The hybrid system's background error covariances (BECs) were a blend of those in 3DVAR and produced by the cycling EnSRF system, and the 3DVAR, EnSRF, and hybrid systems differed almost exclusively by their BECs. New analyses were produced every 6 h between 0000 UTC 1 June and 1800 UTC 14 July 2010 over a domain encompassing the contiguous United States (CONUS) and adjacent areas. Additionally, a control experiment that only assimilated meteorological observations was performed. Each 1800 UTC analysis initialized a 48 h Weather Research and Forecasting with Chemistry model forecast. These forecasts were evaluated with a focus on air quality prediction. The ensemble aerosol spread was generally insufficient, particularly over the western CONUS. However, despite the suboptimal ensemble spread, the hybrid system performed quite well and usually produced the best aerosol forecasts. Additionally, both the 3DVAR- and EnSRF-initialized forecasts typically outperformed the control. These results are encouraging and suggest the resiliency of the hybrid method. Improved aerosol ensembles should translate into even better future hybrid forecasts.

[1] The capability of assimilating surface PM10 (particulate matter with diameters less than 10 µm) observations has been developed within the National Centers for Environmental Prediction Gridpoint Statistical Interpolation three-dimensional variational (3DVAR) data assimilation (DA) system. It provides aerosol analyses for the Goddard Chemistry Aerosol Radiation and Transport aerosol scheme within the Weather Research and Forecasting/Chemistry model. Control and assimilation experiments were performed for June 2011 over China to explore in detail the impact of assimilating surface PM10. In the assimilation experiment, analyses were produced every 6 h to adjust the mass concentrations of different aerosol species. The statistical results from two parallel experiments demonstrate that the assimilation of surface PM10 observations can significantly reduce the uncertainty of initial aerosol fields and effectively improve the subsequent aerosol forecasts for at least 12 h. However, the benefit from the assimilation of PM10 diminishes rapidly with forecast range. Process analysis for PM10 formation indicates that the rapidly diminishing DA impact on aerosol forecasts, especially in early forecast hours, was dominated by vertical mixing with an additional contribution from advection. Both processes were mainly related to unbalanced aerosol fields in the horizontal and vertical after assimilating surface observations. These findings illustrate the importance of adjusting aerosol emission rates and the initial aerosol vertical profile.

Data assimilation is the process by which observations are combined with short-range NWP model output to produce an analysis of the state of the atmosphere at a specified time. Since its inception in the late 1990s, the multiagency Weather Research and Forecasting (WRF) model effort has had a strong data assimilation component, dedicating two working groups to the subject. This article documents the history of the WRF data assimilation effort, and discusses the challenges associated with balancing academic, research, and operational data assimilation requirements in the context of the WRF effort to date. The WRF Model's Community Variational/Ensemble Data Assimilation System (WRFDA) has evolved over the past 10 years, and has resulted in over 30 refereed publications to date, as well as implementation in a wide range of real-time and operational NWP systems. This paper provides an overview of the scientific capabilities of WRFDA, and together with results from sample operation implementations at the U.S. ...

The Weather Research and Forecasting (WRF) model–based variational data assimilation system (WRF-Var) has been extended from three- to four-dimensional variational data assimilation (WRF 4D-Var) to meet the increasing demand for improving initial model states in multiscale numerical simulations and forecasts. The initial goals of this development include operational applications and support to the research community. The formulation of WRF 4D-Var is described in this paper. WRF 4D-Var uses the WRF model as a constraint to impose a dynamic balance on the assimilation. It is shown to implicitly evolve the background error covariance and to produce the flow-dependent nature of the analysis increments. Preliminary results from real-data 4D-Var experiments in a quasi-operational setting are presented and the potential of WRF 4D-Var in research and operational applications are demonstrated. A wider distribution of the system to the research community will further develop its capabilities and to encourage testing under different weather conditions and model configurations.

[1] Total 550 nm aerosol optical depth (AOD) retrievals from Moderate Resolution Imaging Spectroradiometer (MODIS) sensors and surface fine particulate matter (PM2.5) observations were assimilated with the National Centers for Environmental Prediction (NCEP) Gridpoint Statistical Interpolation (GSI) three-dimensional variational (3DVAR) data assimilation (DA) system. Parallel experiments assimilated AOD and surface PM2.5observations both individually and simultaneously. New 3DVAR aerosol analyses were produced every 6 h between 0000 UTC 01 June and 1800 UTC 14 July 2010 over a domain encompassing the continental United States. The analyses initialized Weather Research and Forecasting-Chemistry (WRF-Chem) model forecasts. Assimilating AOD, either alone or in conjunction with PM2.5 observations, produced better AOD forecasts than a control experiment that did not perform DA. Additionally, individual assimilation of both AOD and PM2.5 improved surface PM2.5 forecasts compared to when no DA occurred. However, the best PM2.5 forecasts were produced when both AOD and PM2.5 were assimilated. Considering the goodness of both AOD and PM2.5 forecasts, the results unequivocally show that concurrent DA of PM2.5 and AOD observations produced the best overall forecasts, illustrating how simultaneous DA of different aerosol observations can work synergistically to improve aerosol forecasts.

[1] Assimilation of the Moderate Resolution Imaging Spectroradiometer (MODIS) total aerosol optical depth (AOD) retrieval products (at 550 nm wavelength) from both Terra and Aqua satellites have been developed within the National Centers for Environmental Prediction (NCEP) Gridpoint Statistical Interpolation (GSI) three-dimensional variational (3DVAR) data assimilation system. This newly developed algorithm allows, in a one-step procedure, the analysis of 3-D mass concentration of 14 aerosol variables from the Goddard Chemistry Aerosol Radiation and Transport (GOCART) module. The Community Radiative Transfer Model (CRTM) was extended to calculate AOD using GOCART aerosol variables as input. Both the AOD forward model and corresponding Jacobian model were developed within the CRTM and used in the 3DVAR minimization algorithm to compute the AOD cost function and its gradient with respect to 3-D aerosol mass concentration. The impact of MODIS AOD data assimilation was demonstrated by application to a dust storm from 17 to 24 March 2010 over East Asia. The aerosol analyses initialized Weather Research and Forecasting/Chemistry (WRF/Chem) model forecasts. Results indicate that assimilating MODIS AOD substantially improves aerosol analyses and subsequent forecasts when compared to MODIS AOD, independent AOD observations from the Aerosol Robotic Network (AERONET) and Cloud-Aerosol Lidar with Orthogonal Polarization (CALIOP) instrument, and surface PM10 (particulate matter with diameters less than 10 μm) observations. The newly developed AOD data assimilation system can serve as a tool to improve simulations of dust storms and general air quality analyses and forecasts.

[1] This paper reports the first effort on real-time ensemble predictions of precipitation during the 2000–2002 Mei-Yu seasons (May to June) over the Taiwan area. Six members were included, each using the fifth-generation Pennsylvania State University-National Center for Atmospheric Research Mesoscale Model (MM5) nesting down to 15-km grid size, with different combinations of cumulus and microphysics parameterizations. Rainfall forecasts were evaluated with the equitable threat score (ETS) and bias score (BS). On the basis of verifications on 15-km grid points over three Mei-Yu seasons, it was found that no one member persistently had the least root mean square error of 12–24 hours and 24–36 hours accumulated rainfalls. For rainfall occurrence, most members had better predictions over the northeastern mountainous area, the northwestern coastal plain, the central mountain slope, the southwestern coastal plan, and the southwestern mountainous area. These regions also corresponded to areas of more accumulated rainfalls during three Mei-Yu seasons. An ensemble prediction, using a multiple linear regression (MLR) method which performed a least-square fit between the predicted and observed rainfalls in postseason analysis, had the best ETS and BS skill. The MLR ensemble forecast outperformed the average forecast (for all six members), the average forecasts of cumulus (four-member) and microphysics (three-member) ensembles, and also a high-resolution (5-km) forecast; however, a high-resolution forecast still had better skill for heavy rainfall events. The MLR ensemble forecast, using the weightings determined from previous Mei-Yu seasons, still had similar ETS trend to that with weightings determined by current-year Mei-Yu season, albeit with less skill.