Premium
A noise‐aware combination of dual‐frequency measurements from GPS radio occultation
Author(s) -
Wee TaeKwon,
Kuo YingHwa
Publication year - 2013
Publication title -
journal of geophysical research: atmospheres
Language(s) - English
Resource type - Journals
eISSN - 2169-8996
pISSN - 2169-897X
DOI - 10.1002/2013jd019840
Subject(s) - noise (video) , radio occultation , global positioning system , computer science , dual (grammatical number) , a priori and a posteriori , synthetic data , radio frequency , noise measurement , remote sensing , acoustics , algorithm , artificial intelligence , telecommunications , physics , noise reduction , geography , art , philosophy , literature , epistemology , image (mathematics)
One of the fundamental difficulties that arise when using GPS Radio Occultation (RO) data in exploiting the stratosphere is that the air becomes rarefied with increasing height and accentuates the ionospheric effect and noise contained in the measurement. Customarily, the conventional linear combination (CLC) is used to extract neutral atmospheric components from dual‐frequency (L1 and L2) RO data. The CLC combines and magnifies measurement noises, and thus works well only for those measurements of low noise. Although the L1 data are of considerably higher quality than the L2 data, the CLC does not take this into account and treats both equally; this makes the CLC‐produced data less attractive. The authors propose a new approach, named Noise‐Aware Combination (NAC), which is a generalized combination that factors in the presence of measurement noise. In this NAC method, the L1 and L2 data are each regarded independently, with each contributing to the combination according to its dynamically assessed accuracy. The performance of both the CLC and NAC are tested with two sets of data: one of synthetic data and the other of real data. The tests confirm that the NAC yields significant error reductions when compared to the CLC. While the noise in the CLC‐produced data stands out in high altitudes and compels the data to be blended with the a priori, the NAC relies far less on this blending. The clear advantage of the NAC over the CLC would greatly enhance the value of RO for climate research.