Developments in the correction of radiosonde relative humidity biases
Sharpe, Martin ; Macpherson, Bruce
Currently, the Observation Processing System adjusts radiosonde relative humidity observations before assimilation into the NWP models, since using unadjusted values can result in spurious depletion of analysed cloud cover. The current adjustment scheme is a statistically derived bias relationship applied as a function of observed value, adding moisture to observations that are close to saturation. Fundamentally, this method relies on a relationship between relative humidity and cloud cover, as defined by the model cloud scheme. It uses cloud cover estimates derived from relative humidity profiles to make comparisons with surface reports of cloud, which then allows humidity adjustments to be calculated. Real time adjustments are desirable because their design can make them robust to future changes in measuring equipment and model physics, whereas the statistical scheme requires repeated re-derivation in order to account for these things. This paper describes such a scheme, based on the current statistical approach, in detail and presents the results of tests using the mesoscale versions of the Unified Model. The test results showed a slightly negative overall forecast impact. Further investigation and tuning of the scheme is planned, with the hope of including it in a future Observation Processing System upgrade having obtained positive forecast impacts.
Accès à la notice sur le site du portail documentaire de Météo-France