Predictability of weather and climate

Abstract:

Radio occultation experiments have been simulated for the Arctic region on the basis of the regional atmospheric model HIRHAM4. Irregular structures in the atmosphere produce a violation of the quasi-sphericity in the radio signal propagation and exert a strong influence on the accuracy of atmospheric profiles retrieved by the radio occultation technique. Errors in radio occultation data are spatially localised and associated with gradients in atmospheric structures. Local errors reach 2% in retrieved profiles of refractivity corresponding to an error of 6 K in temperature. Therefore mesoscale variations in atmospheric parameter gradients in a specified region must be taken into account when interpreting radio occultation data. We show, that a correction functional can be developed using the refractivity index field calculated from the regional model in order to improve the radio occultation retrieval of atmospheric parameters. This functional is constructed from instantaneous model outputs, as well as from temporally averaged fields of refractivity using data of the HIRHAM4 model for the Arctic atmosphere. The correction functional derived from monthly averaged data reduced the retrieval errors of refractivity, temperature, and pressure in the troposphere, in particular, temperature retrieval errors are reduced up to 1 K. Application of this kind of functional depends on whether the model used for the construction of the functional is able to simulate the real mesoscale atmospheric structures. © 2004 Elsevier Ltd. All rights reserved.

Abstract:

Shadowing trajectories can play an important role in assessing the reliability of forecasting models, they can also play an important role in providing state estimates for ensemble forecasts. Gradient descent methods provide one approach for obtaining shadowing trajectories, which have been shown to have many useful properties. There remains the important question whether shadowing trajectories can be found in very high-dimensional systems, like weather and climate models. The principle impediment is the need to compute the derivative (or adjoint) of the system dynamics. In this paper we investigate gradient descent methods that use limited derivative information. We demonstrate the methods with an application to a moderately high-dimensional system using no derivative information at all. © 2003 Elsevier B.V. All rights rserved.