Progress in Assimilation of Lightning Data into a Mesoscale NWP Model

Diabatic heating sources, especially latent heat release in deep convective clouds can play an important role in storm development and dynamics. Lack of observations over the Pacific ocean can lead to inadequate initialization of the numerical models and large errors in storm central pressure and rainfall forecasts. Specifying diabatic heating sources in the early hours of the forecast can improve the model's performance. Data from Pacific Lightning Detection Network (PacNet) are used to identify the areas and intensities of convective activity and latent heat release over the Pacific ocean storms.

In a recent experiment, lightning data from PacNet were assimilated into MM5 using a lightning-rainfall-moisture profile relationship. The lightning-rainfall relationship was derived by comparing PacNet lightning data and simultaneous satellite rainfall data. FDDA (Four-Dimensional Data Assimilation) was used to assimilate lightning data into MM5. The model predicted vertical moisture profiles were nudged towards moisture values derived from lightning data. Lightning-rainfall relationship was used to relate lightning rates with moisture profiles. Seven vertical moisture profiles typical for a range of rainfall rates were constructed. The results from this experiment were very promising: in a case study of an extratropical cyclone in the North-East Pacific, 12 hour MM5 control forecast showed 10 mb error in the storm central pressure, whereas the error was reduced to 1 mb when using lightning data assimilation. In another case, the method was able to fix the position of a squall line over Hawaii, which was forecasted 150 km off in the control run.

A second approach is being tested, which is to scale latent heating tendencies at each grid point in MM5, depending on the ratio between model predicted rainfall and rainfall derived from lightning data. In an operational forecast system, latent heat assimilation has some advantages over moisture profile assimilation. Construction of the moisture profiles requires prior knowledge of the temperature, whereas latent heating scaling technique is independent of environmental temperature, making this approach more robust.