Revalidation of TAMDAR data needed Errors found in
- Slides: 8
Revalidation of TAMDAR data needed • Errors found in GLFE co-location tests have been addressed by Air. Dat • Affected Temperature and Relative Humidity • Errors in Winds need to be assessed • Large errors in GLFE aircraft navigation systems precluded adequate assessment • Questions of impact of accuracy of TAMDAR reports from higher speed jet aircraft needs to be assessed • E. g. , dynamic heating effects on Relative Humidity ___________________ Take all those old studies and throw them out! Mark Anderson 4/12/06
Revalidation of TAMDAR data needed • Errors found in GLFE co-location tests have been addressed by Air. Dat • Affected Temperature and Relative Humidity • Errors in Winds need to be assessed • Large errors in GLFE aircraft navigation systems precluded adequate assessment • Questions of impact of accuracy of TAMDAR reports from higher speed jet aircraft needs to be assessed • E. g. , dynamic heating effects on Relative Humidity ___________________ • GLFE experience showed that a variety of validation approaches were needed: • “Dedicated” Radiosonde Co-locations (tight time/space constraints) provide highly accurate results at one location • “Radiosondes of Opportunity” validate and expand representativeness of “Dedicated” co-locations, but at expense of time/space fits • Model comparisons expand time and space coverage, but at the expense of a less precise “truth” and possible introduction of model specific baises.
Comparisons of Sfc and 4 km Wind Bias and Vector RMS for 1, 3, 6, 9 hr Fcsts For Vici and Haskell OK Model forecasts alone do not provide an adequate validation standard Sfc – Green 4 km – Black Sfc-4 km Shear- Red
Comparisons of Sfc and 4 km Wind Bias and Vector RMS for 1, 3, 6, 9 hr Fcsts For Vici and Haskell OK Model forecasts alone do not provide an adequate validation standard Sfc – Green 4 km – Black Sfc-4 km Shear- Red Difference in error between 1 hr old ob and 3 hr RUC forecast Note: Errors in 1 hr RUC forecast > Presistance of previous hourly ob at both sfc and 4 km and at both sites. Difference in error between 1 hr old ob and 3 hr RUC forecast
Comparisons of Sfc and 4 km Wind Bias and Vector RMS for 1, 3, 6, 9 hr Fcsts For Vici and Haskell OK Model forecasts alone do not provide an adequate validation standard Difference in error between 1 hr old ob and 3 hr RUC forecast Difference in bias between 1 hr old ob and 3 hr RUC forecast Note: Errors in 1 hr RUC forecast > Presistance of previous hourly ob at both sfc and 4 km and at both sites. Biases and RMS also change from site to site. Sfc – Green 4 km – Black Sfc-4 km Shear- Red
Model forecasts alone do not provide n adequate validation standard Errors in 1 hr RUC forecast > Presistance of previous hourly ob at both sfc and 4 km. Biases and RMS also change from site to site. In addition, model error (uncertainty) along West Coast greater than in GLFE area Need another Radiosonde co-location validation to provide a ground truth of these relatively large changes in TAMDAR
Need to know route structure of TAMDAR equipped aircraft to assure sufficient co-located observations will be possible Desirable to use airport with minimal congestion to optimize time/space co-location
Need to know route structure of TAMDAR equipped aircraft to assure sufficient co-located observations will be possible Desirable to use airport with minimal congestion to optimize time/space co-location Proposing that NWS (NASA? ) a Radiosonde co-location in Portland after sufficient TAMDAR units have installed and ‘checked out” on Horizon aircraft
