An inverse modelling method with Doppler lidar measurements to estimate the surface sensible heat flux

Traditional point measurements of heat fluxes, e.g. using eddy covariance methods with a sonic anemometer, have a limited source area, making them of limited value in estimating area-averaged heat fluxes in heterogeneous sites, such as cities. We have developed a novel technique that uses an optimal inverse method to retrieve the heat flux from profiles of the vertical velocity variance using mixed layer scaling relationships. The estimated fluxes are representative of a larger source area because the lidar measures variances through the whole boundary layer depth. The method is tested over a surface that is considered homogeneous, such that the source area of the measurement is unimportant and comparison can be performed with measurements from a sonic anemometer. The results are encouraging, although the lidar-estimated heat flux tends to be smaller compared to the sonic-estimated values, particularly in the morning. Sources of error include the sampling error associated with the variance estimate, and the rapid deepening of the boundary layer, particular during the morning, during the averaging period.