Near surface air temperature lapse rates over complex terrain: a WRF based analysis of controlling factors and processes for the central Himalayas

Various environmental processes are strongly controlled by spatio-temporal variations of surface air temperature (hereafter temperature) in complex terrain. However, the usually scarce network of observations in high mountain regions does not allow for an investigation of the relevant micro-meteorological processes that result in complex temperature fields. Climate impact studies often utilize a constant lapse rate of temperature (LRT hereafter) in order to generate spatially distributed temperature data, although it is well acknowledged that LRTs feature a pronounced variability at spatial, seasonal, and diurnal scales. In this study, the Weather Research and Forecasting (WRF) model is used to understand the factors and processes influencing temperature and LRT in the Khumbu and Rolwaling regions of the central Himalayas. A high resolution simulation is performed for one complete year (June 2014-May 2015) in order to capture the entire seasonal cycle. To test the model response to land cover and terrain characteristics, additional simulations with adjusted surface conditions are conducted. Our results demonstrate the capability of WRF to reproduce the processes controlling LRT, although an LRT bias is detected during non-monsoon seasons. The simulated temperature fields feature two LRT minima (i.e. low temperature decrease with elevation) during Nov-Dec and monsoon season, and two LRT maxima (strong temperature decrease with elevation) during the early post-monsoon and pre-monsoon seasons. A steeper LRT (i.e. a rapid decrease of temperature with elevation) is found at high elevations (> 4500 m) while shallower LRT values (i.e. a slower decrease or even increase of temperature with elevation) are apparent at lower elevations. During the pre-monsoon season, high net insolation rates and a reduced latent heat loss from snow free surfaces cause strong sensible heating at low elevations, while the presence of snow at high elevations leads to reduced sensible heating. This strong contrast results in steeper LRT values. Early post-monsoon shows similar characteristics but with a reduced magnitude. The shallow LRT during monsoon season is shown to be caused by the large-scale moisture supply and the associated latent heat release at the Himalayan slopes. This effect is further intensified due to strong up-valley winds which contribute to a well-mixed troposphere. Temperature inversions associated with cold air pooling cause shallow LRT values in lower river valleys during Nov-Dec. The results suggest that the identified processes should be considered for downscaling applications, particularly if distributed temperature fields are required for climate impact investigations.