Assessment of surface layer parameterizations in ARW using micro‐meteorological observations from a tropical station. (5th February 2016)

• Record Type: Journal Article
• Authors: Hari Prasad, K. B. R. R.; Venkata Srinivas, C.; Venkateswara Naidu, C.; Baskaran, R.; Venkatraman, B.
• Is Part Of: Meteorological applications. Volume 23:Number 2(2016)
• Page Start: 191
• Page End: 208
• Publication Date: 2016-02-05
• Subjects: surface layer -- ARW -- seasonal variation -- similarity theory; Meteorology -- Periodicals; Meteorological services -- Periodicals; 551.5
• DOI: 10.1002/met.1545
• Languages: English
• ISSNs: 1350-4827

Abstract

ABSTRACT: The performance of two surface layer parameterizations (MM5‐similarity/Eta‐similarity schemes) in an Advanced Research Weather Research and Forecasting (ARW) mesoscale model is assessed over the tropical site Kalpakkam in southeast India. The ARW is run with four‐nested domains and at a high resolution (1 km) for four 2 day periods covering eight days in each season, 9–17 January, 4–12 April, 4–12 August and 13–21 October 2014 representing winter, summer, the southwest monsoon and the northeast monsoon, respectively. Turbulence observations on winds collected using a fast‐response ultrasonic anemometer are used to estimate the surface layer parameters of friction velocity ( u * ), and Monin–Obukhov length ( L ), and surface fluxes of heat and moisture using the eddy correlation method. Results of simulations indicate seasonal variation in friction velocity, drag co‐efficient, turbulent fluxes of heat and moisture, air temperature, humidity and winds at the study site. Relatively higher u *, drag co‐efficients, turbulent fluxes of heat and moisture are found during summer and the southwest monsoon, followed by the northeast monsoon and winter due to the prevalence of more unstable conditions in these seasons. Although the diurnal cycle of various parameters is well produced by both the schemes, the Eta‐similarity scheme produced higher values of surface layer variables relative to the MM5‐similarity scheme. Although both schemes depict a systematic bias in various surface layer parameters, the predicted values with MM5‐similarity scheme are in better agreement with observational estimates as indicated by the qualitative and quantitative comparisons of results. The differences in the results of the two schemes are attributed to the variation in the empirical stability correction functions and the stability regimes used in the two schemes.