No abstract

Abstract

This case study examined how well downscaling of Community Earth System Model (CESM) data can reproduce climatological conditions relevant for summer (JJA) air quality in Glacier Bay National Park. Climatology was determined from the meteorological results obtained by the Weather Research and Forecasting model inline coupled with chemistry (WRF-chem) when driven with CESM data of 2006-2012. The climatology of this experiment (EXP) was evaluated by climatology from gridded blended sea-wind speeds, CRU data, and 42 surface meteorology sites. The quality relative to known performance was assessed by comparison to climatology determined from WRF-chem control simulations driven with FNL analysis data (CON) in forecast mode. Compared to observations, the thermodynamic and dynamic performances of EXP showed similar shortcomings (dampened diurnal temperature range, overestimation of wind speed over land) as CON. Over water EXP wind-speed climatology JJA bias (simulated minus observed) was −0.7 m/s. With respect to the CRU data EXP biases in JJA 2m temperature, diurnal temperature range, relative humidity and accumulated precipitation were −1.1 K, −4.9 K, 13%, and 110 mm, respectively. The slightly warmer atmosphere in EXP compensated for deficiencies in the cloud schemes leading to better results for the number of wet days and accumulated precipitation than in CON. Downscaling captured known mesoscale responses important for regional climate in a similar way as CON. When using CESM forcing, lateral boundary effects expanded spatially farther into the domain N. Mölders et al. 590 than known for forcing by analysis data. Overall, climatologies obtained from downscaling for Southeast Alaska had similar skill than those derived from forecasts driven by analysis data.

1. A comprehensive decadal evaluation of WRF/Chem over the U.S. using surface and satellite datashows an overall good performance 2. WRF/Chem outperforms WRF in terms of radiative variables due to the inclusion of chemistry feedbacks to climate 3. WRF/Chem outperforms CESM in terms of domain-average performance statistics, and temporal/spatial variations due to a finer grid scale

No abstract