Premium
Investigation of the sensitivity of the ECMWF radiation scheme to input parameters using the adjoint technique
Author(s) -
Janiskov'a Marta,
Morcrette Jeanjacques
Publication year - 2005
Publication title -
quarterly journal of the royal meteorological society
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.744
H-Index - 143
eISSN - 1477-870X
pISSN - 0035-9009
DOI - 10.1256/qj.04.183
Subject(s) - sensitivity (control systems) , parametrization (atmospheric modeling) , jacobian matrix and determinant , environmental science , radiation , cloud cover , meteorology , mathematics , computer science , radiative transfer , cloud computing , physics , quantum mechanics , electronic engineering , engineering , operating system
The potential of the adjoint technique for studying the sensitivity of a physical parametrization scheme to its input parameters is investigated. Compared with the standard evaluation approach, where the sensitivity of all the outputs is obtained by modifying one given input, the adjoint method offers a complementary and very efficient approach for sensitivity studies. The paper focuses on an estimation of the sensitivity of the ECMWF radiation schemes to changes in temperature, humidity, aerosols and cloud properties. The sensitivity of the topof‐atmosphere radiation fluxes in clear‐ and cloudy‐sky conditions, computed globally by the ECMWF model for the different seasons, is discussed in this study. This allows us to investigate spatial and temporal patterns in the sensitivity variations. The evaluation of the Jacobian of radiation fluxes by the adjoint technique provides an insight on the extent to which the various radiation fluxes are sensitive to the meteorological variables and confirms results usually obtained by more computationally intensive traditional methods. In addition, such a sensitivity study gives some indications on the importance and efficiency of particular types of observations in modifying the initial conditions of a model. Copyright © 2005 Royal Meteorological Society.