NRL COAMPS Real-Time Forecast during VOCALS-Regional Experiment

VOCALS Regional Experiment (VOCALS-Rex) is designed to understand the interaction among aerosol, clouds, oceans, and the Andes Mountains; it was conducted in October – November, 2008 off the west coast of Peru and Chile. During VOCALS-Rex, NRL COAMPS was used to provide weather forecast for the experiment. This work is to evaluate the COAMPS forecast with a focus on the cloud-topped marine boundary layer structure using satellite and in-situ measurements taken in the field program. The in-situ measurements include rawindsonde soundings launched from NOAA R/V Ron Brown, soundings from NPS Twin-Otter near the location (72W, 20S) and cross-section flights from NSF C-130; the satellite observations include those from AMSR-E, TMI and SSM/I F13 and F15.

COAMPS is configured to include three nested grids (45km, 15km, 5km) with 45 levels in the vertical; it provides 48-hour forecast twice daily between October 20 – November 29, 2008. For the evaluation and analyses conducted in this study, only the results on the 15km grid are presented.

The comparison of the forecasts with the data shows the followings:

1. The general pattern of the 40-days averaged liquid water path and surface wind speed from the forecasts is consistent with the satellite observations. The forecast cloud diurnal variation is significantly weaker than the observations in both the amplitude and the regression correlation coefficient.

2. Surface wind appears to be a better forecast variable than the liquid water depth, and precipitable water.

3. Predicted potential temperature profiles in general follow those from soundings, both within and above the boundary layer top away from the coast, indicating that the main thermal forcing for the BL is acceptable. The correlation between the forecast and observed water vapor is particularly poor above the boundary layer, which may result from the forecast error in the wind variability.

4. The boundary layer height near the coast is significantly lower than the observations; our budget analysis suggests that it may result from the unusually strong subsidence along the coast in the forecast.

5. The forecast boundary layer height increases toward the west, although its slope is rather gentle away from the coastal region compared with the observations.

Further discussions will be given in the presentation.