Research

 

Understanding the Impacts of Assimilating Satellite-Derived Data on Mesoscale Analyses and Forecasts of Tropical Cyclones

Motivation

The goal is to provide more accurate mesoscale analysis for tropical cyclone (TC) forecast by improving TC structure and its environments via assimilating high resolution satellite data with an Ensemble Kalman Filter.


Model and Data Assimilation System

WRF (The Weather Research & Forecasting Model)

DART (Data Assimilation Research Testbed) from NCAR

Part I - The Assimilation of Atmospheric Motion Vectors (AMVs)

















  1. BulletWith an order of magnitude more AMVs assimilated, the analyses and forecasts that utilize CIMSS

hourly AMV exhibit superior track, intensity and structure.


  1. BulletThe modification to TC structure is subtle with the additional Rapid-Scan AMVs assimilated,

however, forecast track is shifted to the more accurate orientation.




Part II - Understanding the Contributions of Assimilating Subsets of AMVs




























  1. BulletUpper-level and Interior AMVs are most important in maintaining the initial track and MSLP, and later on more accurate forecast track and intensity.


  1. BulletWithholding upper-level and Interior AMVs can also lead to large model uncertainties to estimate storm-relative quantities such as tangential wind and radial wind.


  1. BulletThe pre-mature recurvature found when Rapid-Scan AMVs are assimilated is large associated with mid-level AMVs.

Part II - The Influence of Assimilating AIRS Temperature and Moisture Soundings and AMSR-E Total Precipitable Water

































  1. BulletPreliminary results suggest that the assimilation of AIRS temperature and moisture soundings has subtle improvement over CTL, a parallel experiment in which no AIRS soundings were assimilated. Bias correction to AIRS sounding is necessary to remove significant cold bias in upper troposphere and moist bias in lower atmospheric layers.