CS86 Experiment 1: Aviation weather modeling

Introduction

Aviation traffic on the runway use in an airport depends on the weather. For example, low visibility may cause an airport close, and wind direction dictates the runway use. In this experiment we will investigate any possible relationships among several weather parameters. If we are able to establish relationships among the weather parameters, we may be able to predict aviation traffic based on those weather parameters that affect aviation traffic in an airport.

Problem

In this experiment we will investigate any possible relationships among temperature, dew point, visibility, wind speed, and wind direction. In particular,

1.  the temperature spread defined by the absolute difference between temperature and dew point in F, and visibility will be studied for any possible correlation.

2. Also, the correlation between wind direction and runway will be studied, as well as the correlation between temperature spread and runway use, and the correlation between visibility and runway use.

Hypothesis

The hypothesis to be studied is the existence of a correlation between the temperature spread and the visibility vs non-existence of the correlation.

Apparatus

The FAA DUATS aviation weather system, the ASOS and ATIS weather systems in the local airports such as LGA and JFK will be accessed for retrieving the weather data and the aviation traffic information needed for this experiment.

The weather data and runway information will be analyzed using MathCAD and S-PLUS software available in a personal computer.

Procedure

Step 1.

Access the ASOS and ATIS weather systems in the LAG and JFK airports via publicly accessible telephone lines. At least twenty weather samplings should be conducted at each airport, with each weather sampling separated from the next one by one hour.  On each data sampling, the information to be collected should include: temperature, dew point, visibility, wind speed and direction. Also, date and time the weather information applies, as well as the type "surface observation" should be noted. The data collected under this step of procedure will be referred to as DataSet1.

Step 2.

Analysis

Step 1.

Compute the temperature spread using the data in the DataSet1 as discussed in the document "http://tweety.geol.qc.edu/nsf/demo/bondemo.mcd". It will be desirable to use MathCad to perform this task.

Step 2.

Compute the correlation between the temperature spread and the dew point using DataSet1. This should be done using MathCad. While you are computing the correlation, you should also record the mean, variance, and the co-variance values along your calculation.

Step 3.

Make a scatterd plot for the correlation analysis you do in step 2.

Step 4.

Repeat steps 2 to 3 but focus the analysis (1) on  the correlation between wind direction and runway,  then (2) on the correlation between temperature spread and runway use, and finally (3) on the correlation between visibility and runway use.

Discussion

1. Can you determine whether the stated hypothesis can be accepted or rejected?

2. Can you determine any relationships between the results you obtained from the correlation analysis (done analytically and mathematically), and the scattered plot (done for visual inspection)?

3. Given the airport information distributed in the class, if you see one day many airplanes flying over Queens College, can you make any statement about the weather condition in the locality of Queens College?

4.  As a follow up for (3), what about the another way around that you have the weather information and you want to make statement(s) about airplanes flying over Queens College?

5. What were some problems with the present experiment that should be controlled when the experiment is conducted again? (Hint: consider the recording time and location).