Progress in LWS TR

1
Progress in LWS TRT Research

• June 2004
• Steve Naehr, Frank Toffoletto
• Rice University

2
Data-based analysis of DBx at GEO
Statistical interpolation combines observations
with a background model, to improve the estimated
state globally. The figure shows predicted DBx
at the position of GOES-12, computed from the
Tsyganenko (2001) field model and data from
GOES-10.
3
Estimating error correlations in the Tsyganenko
(2001) B-field model
Statistical interpolation requires error
structure of background model. Here, error
correlations are estimated numerically, by
perturbing model inputs and parameters. Left
contours of DBz error correlation
coefficient, relative to point marked in
equatorial plane.