Retrieval of flow structures in a convective boundary layer using an adjoint model: Identical twin experiments

The applicability of a four-dimensional variational data assimilation (4DVA R) technique to retrieval of microscale turbulent structures in a convectiv e boundary layer is assessed. Two new features are implemented into the exi sting 4DVAR model: a height-dependent eddy viscosity and a surface flux mod el. The identical twin experiments approach is adopted to utilize the model itself to generate 13 instantaneous three-dimensional radial velocity data sets uniformly spanning 5 min. An ideal experiment, using these datasets as the observations, is first tested. After 400 iterations, the resulting cor relation coefficients between retrieved and exact data are 0.99 for velocit y and 0.97 for temperature fields. To emulate the lidar scanning feature, t he 13 three-dimensional datasets are used to construct two volume scan data sets with each horizontal data slice taken from different instantaneous dat asets. Using these data as the input, the correlation coefficients for hori zontal, spanwise, and vertical velocity fluctuations and temperature can st ill reach 0.97, 0.97, 0.94, and 0.72 after 400 iterations. Addition of a su rface flux model improves retrieval quality. Allowing height-dependent eddy viscosity and diffusivity does not improve retrieval quality, whereas doub ling the value of eddy diffusivity improves retrieval quality. Implementati on of temporal and spatial smoothness penalty functions significantly impro ves retrieval quality in the presence of various sources of error.