Classification and Regression Tree (CART) Calculator

Introduction : This calculator is a working version of the decision diagram represented by the chart below. This particular decision tree, created in 1988 for the Los Angeles (California) valley, was created using these variables:

  1. the temperature (in degrees Celsius) at the 850-mb pressure level
  2. the temperature (in degrees Celsius) at the 900-mb pressure level
  3. the length of the day (in hours)
  4. the top of two inversion layers (in degrees Celsius)
  5. the direction of the west winds at El Toro Marine Air Base
  6. the direction of the winds at Los Angeles International Airport (LAX)

This decision tree was created from a database of 1,096 observation days, looking at the concentration of ozone (O3) in parts per billion (ppb).

Directions:

Suppose you wish to make a forecast for a particular day. All you know for that day is that the 850-mb temperature is expected to be 16.0 degrees Celsius. Enter that value into the CART calculator. Since you do not know any of the other values, enter zeros or dial in "unknown" for the winds. Click on the "Evaluate" button, and you should have a forecast of 130 ppb, wiht a standard deviation of 47. The calculator also shows that 374 out of 1096 34%) of the observed days had these conditions.

Classification and Regression Tree (CART) Calculator

Input Values:

850-mb temperature (degrees C)
900-mb temperature (degrees C)
Day length (hours)
Morning wind direction at Los Angeles International (LAX)
Top of two inversion layers (degrees C)
West morning winds at El Toro, CA (NZJ)

Results:

Number of observed days with these conditions
Mean Ozone concentration (parts per billion, ppb)
Standard deviation (S.D.)