Sensitivity of NOGAPS Forecasts to UAS-like Observations

Abstract

Abstract : An adjoint-based method is used to calculate the impact of in situ upper-air observations from a data-austere region of high meteorological variability (Almaty, Kazakhstan) on short-range forecast error in the Navy Operational Global Atmospheric Prediction System (NOGAPS). During the May 2006 ? July 2007 study period, Almaty Aircraft Meteorological Data Relay (AMDAR) ascent and radiosonde observations assimilated at 00 UTC have large beneficial impacts on forecast error reduction when compared to average global AMDAR ascent and radiosonde observation impacts. For Almaty, the average impact of an AMDAR ascent observation is more than twice as beneficial as that of a radiosonde observation in the reduction of forecast error in the global domain. The large beneficial impact of Almaty AMDAR ascent observations offers great promise for the beneficial utilization of weather data from unmanned aircraft systems (UAS) deployed in similar environments. Currently, the long-endurance medium-altitude Predator flown by the USAF and others is the most suitable UAS platform available for AMDAR-like surface to mid troposphere atmospheric profiling. The acquisition of both test and in-theater Predator data and the concurrent examination of how such data impacts the accuracy of short-range forecasts in data-sparse regions are ongoing at NRL.