Accession Number:



Adaptive Observations in Fastex IOP-18: Data Impact and Synoptic Interpretation.

Descriptive Note:

Corporate Author:


Report Date:


Pagination or Media Count:



An issue of major importance in numerical weather prediction is the requirement to provide models with accurate initial conditions. The Fronts and Atlantic Storm Track Experiment FASTEX provides an opportunity to examine the impact of special observational data in 1-3 day forecasts of north Atlantic cyclones. An overview of FASTEX objectives and observational resources is contained in July et al 1997. The FASTEX field phase January - February 1997 provided the first test of so-called adaptive observation techniques, in which objective guidance from numerical models was used to target observational resources e.g., dropsonde aircraft to specific areas. In this study we describe the impact of targeted aircraft dropsonde, and GOES-8 wind data on a 24 hr forecast of the cyclone in FASTEX IOP-18. It is well established that numerical forecasts are much more sensitive to initial condition changes in certain locations than in others. In sensitive locations, changes to initial conditions including changes which night result from new observational data can have critical effects on the development of a forecast feature, such as a cyclone, or on measures of forecast error. Adjoint methods can be used to determine sensitivity patterns related to error growth in mid-latitude forecast situations, for example, see Rabier et al. 1996. Both dry and moist adjoint sensitivity studies suggest that mid-latitude synoptic-scale cyclones on the 1-3 day range have maximum sensitivity to temperature and wind in the mid-lower troposphere roughly 400-800 mb, in relatively localized areas of the upstream baroclinic storm track also see Langland et al. 1995. The problem of adaptive observations does not center on identityand correcting the largest analysis errors, but rather on correcting analysis error in critical locations, where any initial error can amplify with very large growth rates.

Subject Categories:

  • Theoretical Mathematics
  • Meteorology

Distribution Statement: