In this paper we discuss the utility of field observations for model development from an Arctic standpoint. Models have to be true to reality as far as is possible, and when it comes to developing model physics, which is crucial for model climate and hence for model performance on all time scales, there are many processes that can only be studied with the dedicated detailed observations that comes from field projects. We argue that field observations can be of use for model developments in three different ways: 1) To discover phenomena and aspects of processes that have not been covered previously; 2) To explore process or variable relationships that are needed either to develop new parameterizations, to evaluate assumptions that lie behind all parameterisations or to determine closure constants in parameterisations; 3) To evaluate models either straightforward, where regular observations are unavailable, or process oriented which is the method we encourage since it is the only means by which compensating errors can be avoided and random errors separated from systematic errors.
BT - ECMWF-WWRP/THORPEX workshop on polar prediction C1 - Events CY - ECMWF DA - 06/2013 LA - eng N2 -In this paper we discuss the utility of field observations for model development from an Arctic standpoint. Models have to be true to reality as far as is possible, and when it comes to developing model physics, which is crucial for model climate and hence for model performance on all time scales, there are many processes that can only be studied with the dedicated detailed observations that comes from field projects. We argue that field observations can be of use for model developments in three different ways: 1) To discover phenomena and aspects of processes that have not been covered previously; 2) To explore process or variable relationships that are needed either to develop new parameterizations, to evaluate assumptions that lie behind all parameterisations or to determine closure constants in parameterisations; 3) To evaluate models either straightforward, where regular observations are unavailable, or process oriented which is the method we encourage since it is the only means by which compensating errors can be avoided and random errors separated from systematic errors.
PB - ECMWF PP - ECMWF PY - 2013 T2 - ECMWF-WWRP/THORPEX workshop on polar prediction T3 - ECMWF-WWRP/THORPEX Workshop TI - Why modellers should care about field projects ER -