Weather forecasting in Alaska is extremely challenging due to the rugged and topographically-diverse terrain and a significant lack of observations. The upper-air observations, or radiosondes, are subject to large space-time coverage gaps, which deteriorate to the west, thus making forecast adjustments based on observed upstream profiles more difficult. Additionally, the sparse data in this region has long plagued the success of numerical weather prediction in Alaska. The complex mountainous terrain and land-sea interface warrant specific tuning of model physics and parameterization to properly utilize the few available observations.

To mitigate the issue, the Alaska-based airline PenAir began equipping a fleet of 10 Saab 340s with Tropospheric Airborne Meteorological Data Reporting (TAMDAR) sensors in late summer 2007, and was fully deployed, calibrated, and reporting by the beginning of 2008. The sensor measures humidity, pressure, temperature, winds aloft, icing, and turbulence, along with the corresponding location, time, and altitude from built-in GPS. These observations are transmitted in real-time to a ground-based network operations center via a global satellite network.

Data-denial studies are carried out over this region using several data assimilations techniques into the NCAR Advanced Research WRF (ARW) and the MM5. Parallel 48 to 72-h experimental (control) simulations that include (withhold) the PenAir TAMDAR data are conducted. Case-specific and time-averaged forecast skill statistics, verified against conventional observing platforms (e.g., RAOBs, ASOS, etc.), are compiled and analyzed for various domains and model configurations in, and downstream of, Alaska. These cases are then compared to the operational handling of the weather event to identify areas of needed improvement, as well as model biases and trends.

The initial objectives of this study are to (i) quantify any impacts that TAMDAR data may have on high-resolution short-range mesoscale forecasts over, and downstream, of Alaska, as well as (ii) monitor the accuracy, tuning, and health of the PenAir-based TAMDAR observing system. An additional objective, which is an inherent byproduct of the study, is to fine-tune the data assimilation in this unique region with minimal observation competition. The ongoing studies and more complete results will be presented at the time of the conference.