Which Weather Forecaster?

Overview

Objectives/Goals Weather is an important part of our everyday lives. My science fair project analyzes the accuracy of seven different San Francisco Bay Area weather forecasters: weather.com, ktvu.com, kron.com, abclocal.com, nbc11.com, cbs5.com, and sfgate.com. Methods/Materials I collected sky formation, precipitation, and high and low temperature forecasts (one day and two day out forecasts) and actual occurrences for 47 consecutive days in the fall of 2006. In the middle of my data collection, I dropped cbs5.com for lack of detailed forecasts. To analyze the data, I entered the data points into an Excel database to compare the forecasts with the actual occurrences. Forecasts are given using qualitative descriptions, but I needed quantitative measures which would be meaningful to weather consumers. For each weather metric, I created conversion charts using embedded IF statements in my Excel database to automate the data translation to a 100% accuracy scale. Results For combined forecast accuracy, nbc11.com was the most accurate (87.7 %) followed closely by ktvu.com (86.2%). Forecasters were most accurate when forecasting precipitation (92 % accurate), followed by temperature (86% accurate and about 2 degrees off), and lastly sky (73% accurate). Conclusions/Discussion Although forecasting the weather is a challenging task, there is neither a significant difference between the accuracy of each forecaster nor the difference between one day and two day out forecasts.

Competition history

• CSEF 2007 Earth & Planetary Sciences · Entry J0711

Resources

• CSEF project summary ↗

Related projects

Closest projects by meaning, across every fair and year in the corpus.

Browse more like this

• All Earth and Environmental Sciences projects
• CSEF Earth and Environmental Sciences projects
• Every CSEF 2007 project