An objective nowcasting tool that optimizes the impact of GOES-Derive Product Imagery in very-short-range forecasts and nowcasts

Many future instruments (e.g., Wind Profilers networks, automated aircraft reports and the Hyperspectal Environmental Sounder planned for GOES-R) have the capability of resolving atmospheric features beyond today's capabilities in both time and space. Although these data are expected to generate improvements in numerical forecast guidance out to 48 hours and beyond, a greater benefit from these high-time-frequency and detailed data sources may come from their use in real time objective nowcasting systems designed to assist forecasters with identifying rapidly developing, small-scale extreme weather events.

These nowcasting systems will need to detect and retain extreme variations in the atmosphere, incorporate large volumes of high-resolution asynoptic data from satellites and other high-resolution systems, and be computationally very efficient. Accomplishing this will require numerical approaches and techniques that are notably different from those used in numerical weather prediction where the forecast objectives cover longer time periods. The nowcasting systems will need to place an emphasis on retaining the accuracy of individual observations and preserving the large gradients seen in these data through time. Speed, however, will be of the essence, since in many cases the detailed information provided in the observations is extremely perishable.

The basis for a new approach to objective nowcasting is presented that uses LaGrangian techniques to optimize the impact and retention of information provided by multiple observing systems. The system is designed to detect extreme variations in atmospheric parameters and preserve vertical and horizontal gradients observed in the various data fields. Analytical tests of such an approach have been performed to determine the ability of the method to retain gradients and extremes in meteorological fields. These tests show that the technique is extremely computationally efficient (since the inertial advective terms no longer dictate that the model time step must be a function of grid spacing), is able to retain sharp gradients and observed maxima and minima, and has the capability of providing timely updates to forecast guidance provided by operational forecast models. Tests of the system using idealized jet streaks as initial conditions have provided new understandings of regimes where turbulent overturning (CAT) is likely and how very narrow dry bands form in water vapor imagery.

Real data tests are currently being conducted at CIMSS - with the goals of identifying details of the environments associated with the onset of significant weather events several hours in advance. The tests use full resolution (10 km) derived layer moisture products from the GOES-10/12 sounders to update and enhance operational RUC forecasts. Initial tests are focusing on the use of multi-layer GOES-Derived Image Product (DPI) moisture data, with the long-term goal of providing a basis for using GOES-R and NPOES data when they arrive. In order to show consistency for operational forecasters between observations and nowcast products, results of the DPI nowcast tests are presented in the form of forecast satellite images. Examples will demonstrate the ability of the LaGrangian system to capture and retain details (maxima, minima and extreme gradients) that are important to the development of convective instability 3-6 hours into the future, even after the IR observations themselves may no longer be available in the areas of severe weather due to cloud development.