You are here

Local Interpretation of Precipitation Projections: Great Lakes

TitleLocal Interpretation of Precipitation Projections: Great Lakes
Publication TypeWeb Resource (web services, tools, etc...)
AuthorsRood, Richard
Citation Key93
Community Notes

Local Interpretation of Precipitation Projections: Great Lakes

This document is on the interpretation of the uncertainty of local precipitation information from climate projections.  This interpretation is to assist planners in the use of this information in planning and resource management. 

The steps in the process are:

  • What are the observations of what has happened?
    • Local?
    • Regional?
    • Continental?
    • Seasonal?
  • Do the climate models represent the underlying mechanisms that determine a locality’s weather?
    • Middle-latitude cyclones
    • Mesoscale convective processes
    • Boundary layer transport
    • Local effects
      • Lake
      • Urban
      • Agricultural
  • Is there consistency between the modeled transport of water and the representation of precipitation?
    • Characterize the ability of the climate model to represent the weather mechanisms responsible for transport of water vapor to a location.
    • Characterize whether or not the model converts the transported moisture in precipitation in a way consistent with observations. 
  • Object-based interpretation:
    • Treat weather events as “objects,” that is, coherent systems with an associated set of measurable parameters 
    • Does the the model represent objects (mechanisms) that are similar to observed objects (mechanisms)?
      • Yes, then model is providing potentially usable information. 
      • No, model is only a generic indiciator of background states.



Some Regionally Relevant Analysis

Currently decadal and century long climate projections are made with global climate models.  In 2011 the state-of-the-art are coupled models of the atmosphere, ocean, land, and cryosphere.  The cryosphere represents ice on land and sea.  All of these models work by dividing the Earth into discrete pieces or cells.  For a global model, the size of the cells in the atmosphere is approximately 100 kilometers in the east-west and north-south directions.   In a precise way, this grid cell size is NOT the same as resolution.  That is, it takes several grid cells to represent a feature; hence, the resolution is, often, formally 10 times coarser.  This nuance on the meaning of resolution is true for motions in the atmosphere and ocean.  On land, the grid cell size has a stronger relation to resolution.

Because planners and resource managers want information on spatial scales much smaller than the grid size of the climate model used to make projections, the information from the climate models is "downscaled" to smaller grids.   The details of downscaling are not discussed in this article.  What is discussed are some regionally relevant implications of the resolution of the climate models.

  1. In climate models the Great Lakes are not resolved.  The Lakes are, at best, blobs, and they are often represented as a wet (saturated) land type.
  2. In observations, precipitation in the summertime is often associated with convective (thunder) storms that are far smaller than the grid scale.  Therefore, summertime preciptation is not well represented.
  3. There are strong local effects of the Lakes that are simply not represented in the global climate models
    1. Lake Breezes
    2. Lake-effect Snow
    3. Local impact on large-scale weather systems

Downscaling methods approximate local effects through a variety of techniques.  Downscaling methods start from a global climate model.  If the global model has NO representation of a particular phenomonon, then the downscaling technique does not start from an initial guess that is climatically realistic.  Hence the uncertainty is very high, and expert analysis must be part of the interpretation.