Validating Air Force Weather Satellite Retrieved 3D Cloud Products against Independent Ground and Space-Based Assets

Air Force Weather (AFW) has documented requirements for global cloud analyses and forecasts to support DoD missions around the world.  Cloud analyses are constructed using passive cloud detection algorithms from 17 different near real time satellite sources.  The algorithms are run on individual satellite transmissions at native satellite resolution in near real time.  These native resolution products are then used to construct an hourly global merge on a 24km grid.  AFW has also recently started creation of a time-delayed global cloud reanalysis to produce a ‘best possible’ analysis for climatology and verification purposes. Cloud forecasts include global short-range cloud forecasts created using advection techniques as well as statistically post-processed cloud forecast products derived from various global and regional numerical weather forecast models.  The result is a mix of cloud products covering different spatial and temporal resolutions with varying latency requirements.

AFW has started to aggressively benchmark the performance of their current capabilities.  Cloud information collected from so called ‘active’ sensors on the ground at the DOE-ARM sites and from space by such instruments as CloudSat, CALIPSO and CATS are being utilized to characterize the performance of AFW products derived largely by passive means.  The goal is to understand the performance of the 3D cloud analysis and forecast products of today to help shape the requirements and standards for a future Numerical Weather Model driven cloud analysis and forecast system driven by advanced 4DVAR techniques.

This presentation will present selected results from these benchmarking efforts and highlight insights and observations between passively and actively derived observations and the impacts of varying spatial and temporal depictions of clouds.