|
Home
Summaries of Ozone Seasons
Forecasting Procedures
Forecasting Tools
Forecasting Performance
Spare The Air Program
Back to STI Home Page
|
Forecasting Tools
Many tools exist for predicting ozone concentrations. Some methods are simple, yet not very accurate; other methods are more complicated and more accurate. We used a number of different tools/techniques to forecast ozone concentrations for the Sacramento region. Forecasters combine the results of these tools to create a consensus forecast, which tends to be more accurate than the results from a single tool since one tool’s strengths balance another’s limitations. The six tools/techniques used most often are:
 |
Climatology
|
|
Criteria
|
|
CART (Classification and Regression Tree)
|
|
Regression Equations
|
|
LSOFT (Long-range Sacramento Ozone Forcasting Tool)
|
|
HYSPLIT (Hybrid Single-Particle Lagrangian Integrated Trajectory)
|
|
| Average number of days with high ozone in the Sacramento region |
|
Max 1-hr
Conc (ppm)
|
Month
|
|
May
|
June
|
July
|
Aug.
|
Sept.
|
Oct.
|
|
0.10
|
3
|
5
|
9
|
8
|
8
|
4
|
|
0.11
|
1
|
4
|
6
|
6
|
5
|
3
|
|
0.12
|
1
|
2
|
4
|
3
|
2
|
2
|
|
0.13
|
1
|
1
|
3
|
2
|
1
|
1
|
|
0.14
|
0
|
1
|
2
|
1
|
1
|
0
|
|
0.15
|
0
|
0
|
1
|
1
|
0
|
0
|
Climatology
Climatology, the study of average and extreme weather conditions, also applies to ozone forecasting. Although not very useful as a predictive tool alone, forecasters use this information to guide and bound their forecasts. Using a table like the one shown here, a forecaster can determine that the occurrence of several days of ozone concentrations above 0.15 ppm in the Sacramento region in June is highly unlikely.
Criteria
The Criteria method uses threshold values (criteria) of meteorological or air quality variables to forecast ozone concentrations. This technique is based on the fact that specific values of certain meteorological and air quality variables are associated with high ozone concentrations. For example, high ozone concentrations are often correlated with hot temperatures and, thus, forecasted hot temperatures indicate a potential for high ozone concentrations. As more criteria are met, the probability of measuring high maximum ozone concentrations increases. The advantage of this tool is that it is easy to develop and operate, but it is not very accurate or well-suited for predicting precise ozone concentrations.
|
Forecasting Criteria for Sacramento, California
|
|
Parameter
|
Criteria
|
|
Maximum surface temperature
|
>95°F
|
|
San Francisco 900 mb temperature
|
>25°C
|
|
San Francisco 850 mb wind speed
|
£5 m/s
|
|
San Francisco 1000-500 mb thickness
|
³576 dm
|
|
San Francisco 500 mb height
|
³582 dm
|
|
Sacramento 5000 ft temperature from 0600 PDT aircraft sounding
|
³70°F
|
|
San Francisco-to-Reno pressure difference at 0600 PDT
|
between –3 and 0 mb
|
|
Travis, CA average wind speed from 0600 to 1100 PDT
|
£5 m/s
|
|
Maximum NO concentration at Del Paso Manor or North Highlands monitoring sites between 0600 and 0900 PDT
|
³40 ppb
|
|
Sutter Buttes’ ozone at 0700 PDT
|
³60 ppb
|
|
CART (Classification and Regression Tree)
CART is a statistical procedure that results in the development of a decision tree to predict ozone concentration categories based on meteorological and air quality variables that are highly correlated with ozone. CART predicts the maximum Air Quality Index (AQI) categories for the current- and next-day periods.
To use the CART decision tree, forecasters answer Yes/No questions at each decision point. Eventually, a forecast is reached that shows the probability of observing the five AQI categories (Good, Moderate, Unhealthy for Sensitive Groups, Unhealthy, and Very Unhealthy).
|
Go to Decision Tree
|
|
Go to Regression Equation
|
Regression Equations
Regression is a statistical technique for describing a relationship among variables. For ozone forecasting the relationship is between future ozone concentrations and other variables (e.g., temperature, wind speed, etc.). Regression equations were developed for the Sacramento region to predict the 1-hr and 8-hr peak ozone concentrations for the current- and next-day periods in each of the four air districts.
|
|
LSOFT (Long-range Sacramento Ozone Forecasting Tool)
The Long-Range Sacramento Ozone Forecasting Tool (LSOFT) is software that uses meteorological model forecast outputs as inputs and uses a combination of CART and regression equation techniques to predict the maximum AQI for the current day and the next five days. LSOFT output provides information that helps SMAQMD staff anticipate and plan for Spare The Air days up to five days in advance.
|
Go to LSOFT Example |
|
Sonoma Technology, Inc.