Data Assimilation & Prediction: Major Projects
The Use of Scatterometer Data in Global Numerical Weather Prediction
Sponsor: NASA Jet Propulsion Laboratory
AER
is a team member of the NASA Scatterometer (NSCAT) Science Team,
and has proposed to be a part of NSCAT's successor, QuikSCAT:
We are focusing on algorithm development for scatterometer data retrieval of ocean surface wind, direction ambiguity removal, fusion with other sensors, and assimilation by numerical weather prediction models.
Assimilation
methods are tested using ERS-1,2 and NSCAT scatterometer data.
We use variational analysis methods to combine scatterometer data with data from many other sources. Variational techniques constrain the wind retrieval algorithm to insure smoothness and dynamic consistancy. We are also involved in modeling scatterometer measurement error and scatterometer error correlation since these are measures of confidence and may be used to weight the data during assimilation.
An example of retrieving winds from NSCAT scatteromenter data is shown below for the case of Hurricane Isidore (Central Atlantic hurricane; Sep 1996). The first image shows ambiguous winds from JPL processing. Red wind barbs represent the first ranked ambiguity from a median filter. "X" marks the observed location of Isidore's center at the time of NSCAT's overpass.
Note that incorrect ambiguities were chosen to the east, south and west of Isidore by the median filter. These errors are most likely due to errors in the background analysis used to initialize the median filter.
After addressing the errors in the background field and performing a variational analysis assimilating NSCAT sigma0 data, the following ambiguities were chosen (i.e., those closest to our analysis).
These winds are in better agreement with Isidore's observed position, and represent an improvement in ambiguity removal. This result shows that other information, when used to remove ambiguity in scatterometer winds, must be accurate or inappropriate ambiguities may be selected.
Links:
| |
|
Information
Major Projects
Publications