LAUSR

WRF simulations with different selections of sub-grid orographic drag over the Tibetan Plateau have been evaluated with observation and ERA-Interim reanalysis. Results show that the sub-grid orographic drag schemes, especially the turbulent orographic form drag (TOFD) scheme, efficiently reduce the 10-m wind speed bias and RMS error with respect to station measurements. With the combination of gravity wave, flow blocking and TOFD schemes, wind speed is simulated more realistically than with the individual schemes only. Improvements are also seen in the 2-m air temperature and surface pressure. The gravity wave drag, flow blocking drag and TOFD schemes combined have the smallest station mean bias (-2.05 °C in 2-m air temperature and 1.27 hPa in surface pressure) and RMS error (3.59 °C in 2-m air temperature and 2.37 hPa in surface pressure). Meanwhile, the TOFD scheme contributes more to the improvements than the gravity wave drag and flow blocking schemes. The improvements are more pronounced at low levels of the atmosphere than at high levels due to the stronger drag enhancement on the low-level flow. The reduced near surface cold bias and high pressure bias over the Tibetan Plateau are the result of changes in the low-level wind components associated with the geostrophic balance. The enhanced drag directly leads to weakened westerlies but also enhances the a-geostrophic flow in this case reducing (enhancing) the northerlies (southerlies), which bring more warm air across the Himalaya Mountain ranges from South Asia (bring less cold air from the North) to the interior Tibetan Plateau.