One approach of linking the study of weather and climate is through the path of weather forecasting and developing ways to apply forecast-assimilation systems to climate models. Studies of this type are often focused on how, for example, the climate model diverges from the weather forecast over a short period of time. With such an approach it is possible to define certain deficiencies, but it is often difficult to understand cause and effect. Further, these deficiencies are not absent from the weather models and are often entangled with errors in the assimilation and inconsistencies between physical parameterizations and well-specified, predicted prognostic variables. The approach developed here takes a different path. Rather than focusing on weather forecasting and predictability, it is recognized that certain phenomena in climate models, like nature, are the accumulated impact of regional and local weather-scale systems. These phenomena may be well represented in climate models, or they may be poorly represented, and hence, are identified as sources of bias. If the relevant regional and local processes can be identified, it is possible to define quasi-isolated dynamical systems which might be productively studied to isolate and remedy a class of bias in the climate model. As an example consider the precipitation over the continental United States. In the western half of the United States during winter, synoptic-scale storms propagating from the Pacific and releasing their moisture as rain and snow through large-scale interaction with topography provides a reasonable construct for studying the model representation. Alternatively, in the summer in the central United States, the moisture is provided, often, by a boundary level jet, and convective precipitation is organized in mesoscale structures. In both cases, subsets of the model elements can in principle be isolated for examination. Such an approach has been applied to the examination of tropical and continental precipitation biases as well as to investigating biased presentations of North Atlantic sea ice. The benefits of the approach are numerous. First, we have had some success in identifying cause and effect, and hence, correction of the bias. In the case where the bias is not remedied, it is possible to place it in the context of the global climate – is it important or not? Further, taking such a process-based approach allows information from different fields, e.g. weather modeling and mesoscale modeling, to be incorporated effectively into climate model development. Finally, it allows the determination of which local and regional mechanisms must be represented in climate models; it is how these mechanisms change with time that will provide robust information for adaptation to climate change.