Zhen Zeng - Wide Scientific Applications of GPS Radio Occultation (RO) Observations
The COSMIC (Constellation Observing System for Meteorology, Ionosphere and Climate) mission is another Global Positioning System (GPS) radio occultation (RO) collaborative project of the United States and Taiwan . COSMIC includes six microsatellites, which were launched on Apr. 15, 2006 . The main purpose of this project is to provide ~2500 real-time soundings per day with homogenous global coverage (see Figure 1, which shows the distributions of COSMIC soundings per day (green) compared to existing radiosonde sites (red)), which can potentially have major impacts on weather, climate and space weather research and forecasting. The science behind COSMIC originates with the RO technique which was used to study other planetary atmospheres since the 1960s. This technique works by measuring the amplitudes and phase delays of radio signals from GPS satellites as they are occulted by the Earth's atmosphere, and then retrieving atmospheric profiles of bending angle, refractivity, temperature, pressure and water vapor, as well as electron density in the ionosphere. (COSMIC website: www.cosmic.ucar.edu)
My research at UCAR focuses on the application of GPS RO data to meteorology, climate and ionosphere research.
As to the neutral atmosphere application, I have demonstrated the feasibility of using RO data for tidal studies with my analysis of several-years of CHAMP (a German RO mission) observations. Comparisons between the migrating diurnal tide (i.e., the 24-hour tide that propagates westward with the motion of the Sun) derived from CHAMP RO temperatures and its variation in the tropics between 10 and 30 km with those simulated those simulated by the Global-Scale Wave Model-Version 2002 and the extended Canadian Middle Atmosphere Model showed consistent model and measurement behavior. More importantly, the CHAMP results motivate the follow-on effort to resolve global migrating and non-migrating (i.e., non-Sun-synchronous) tides in the lower and middle atmosphere, as well as their small-scale variations from COSMIC. Such studies were previously impossible because older RO observations had limited local time coverage. Once the six COSMIC satellites are full deployed, COSMIC will provide continuous measurements at all local times.
Because the fundamental RO observation is a measurement of time, it is well suited for climate monitoring with its long-term stability. Though it is still too early to look for a climate signal from the past and ongoing RO observations, they are also very useful to access the quality of the observations from other sensors like the Microwave Sounding Unit (MSU) and advanced MSU (AMSU) that fly on different NOAA satellites. In collaboration with other scientists at NCAR, I converted CHAMP temperatures between 8 and 25 km to equivalent MSU/AMSU lower stratospheric brightness temperature (Tb) and compared these to the MSU/AMSU Tb derived from both Remote Sensing System Inc. and University of Alabama at Huntsville . It is found that the MSU/AMSU retrievals in the Antarctic LS in winter were biased high relative to the GPS RO measurements.
The GPS RO technique also provides a new opportunity to investigate the structure and variation of the ionosphere, and to evaluate and improve numerical ionospheric models. I compared the values of maximum F-region density (NmF2) from the Global Assimilation Ionospheric Model (GAIM) with the early COSMIC observations, and has found good agreement in both the magnitude and distribution of NmF2. Improvement to ionospheric RO retrievals remains a major challenge for the RO technique. The traditional retrieval method assumes local spherical symmetry of the ionosphere, which is far from reality because of the existence of significant ionospheric horizontal gradients plus the length of the ray path. I am attempting to find a better way to solve the problem through assimilating the GPS slant total electron content (including the ionospheric horizontal gradient information) but not vertical electron density observations using an artificial neural network (ANN) method. Preliminary test results demonstrate the feasibility of this scheme. The next step is to test the extent of possible correction as a function of temporal resolution.
ASP Spotlight June 2007
For more ASP spotlights click here http://www.asp.ucar.edu/spotlight/archive.jsp