CHIPS is the Coupled Hurricane Intensity Prediction System (not to be confused with SHIPS, the Statistical Hurricane Intensity Prediction Scheme); it’s a coupled atmosphere-ocean model that predicts intensities based on environmental variables only.  And here’s a nice paper that Kerry (and others) wrote about it.  Some notes:

  • predictions of hurricane tracks have gotten much better over the last 30 years, while predictions of intensity have had less improvement, though the most successful intensity prediction schemes have been statistical ones (though SHIPS has its problems too, is pretty bad for secondary eyewalls)
  • changes in storm intensity (and formation of secondary eyewalls) are due to a combination of environmental variables and internal processes.  Molinari, Vollaro and Nong, Emanuel papers suggest that formation of a secondary eyewall is driven by external processes, while its evolution is affected more by processes internal to the storm.
  • (SEs associated with significant intensity fluctuations)
  • To test the prediction that intensity changes to storm, Kerry and authors wrote a model for hurricane intensity in which the intensity depends only on external variables (CHIPS!).  CHIPS takes interaction with the ocean into account, since hurricanes cool SSTs as they pass and this can be a negative feedback process (see Schade, Emanuel).
  • CHIPS assumes axisymmetric storm with hydrostatic and gradient wind balance; that the vortex is neutrally stable to slantwise convection, in which the temperature follows a moist adiabatic lapse rate along surfaces of constant angular momentum.  (Can I use this..?)
  • Vertical structure determined by boundary layer moist entropy and vorticity at tropopause; water vapor represented by boundary layer moist entropy and middle troposphere layer (seriously where is this boundary layer info?)
  • Model variables are in potential radius (R) coordinates; means that the center of the storm, near the eye, is really well-resolved (is important because of the pressure gradient etc there), whereas less vital parts of the storm from the edges aren’t as well-resolved.  The potential radius is proportional to the square root of the absolute angular momentum/mass and is defined in terms o the Coriolis parameter, the physical radius r, and the azimuthal velocity V as

  • The average resolution (in R coordinates) is 20km, but can be as small as 1-2 km near the eye of the storm.
  • Their SST varies with time and radius to reflect the air-sea coupling with the storm.  (Apparently there’s a 1-D ocean model…should I use it…mainly seems to depend on entrainment-type processes like for final project, see Schade paper.)
  • Apparently shear can suppress the air-ocean interaction.
  • CHIPS runs really fast, but its axisymmetry means that environmental vertical wind shear has to be parametrized, which is a little unfortunate given how important wind shear is (see Tang, Emanuel ventilation paper).  But there is a parametrization scheme that looked at observed intensity versus the model ones and used a multiple regression algorithm to relate other environmental variables to intensity and shear (?), but they end up ventilating the storm at middle levels.
  • Models were initialized with real or forecasted storm tracks with some environmental info; used NCEP SSTs and atmospheric temperature info from 0000 UTC near the beginning of each storm’s life to calculate the potential intensity.
  • Model’s best in low shear or in areas where ocean mixed layer doesn’t depart too far from climatology.  More sensitive to initialization conditions when shear is present.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s