Civil & Environmental Engineering Data for
STAT 451CM

On this page, we intend to post data sets related to civil and environmental engineering for the new Statistics course (STAT 451CM) designed for CE and ME students.

Geotechnical Data
Data provided by: Dr. Trevor Smith
We conducted direct shear tests to measure phi on sands from the coast retrieved from six sites. Phi governs all slope stability and foundation factors of safety and is therefore VERY important. Possible questions around phi:

1. Is there any relationship to unit weight for these similar Dune sand deposits, either at a site or overall?
2. Any relationship to depth for these sands in a deposit or overall?
3. What are the probability density function shapes most likely to fit, normal or log normal, or ?
4. Characterize the statistical gain from, say, double the number of test results.
5. What is the required number of tests to be run to generate certain confidence level in phi?

Water Resources Data
Data provided by: Dr. Scott Wells

The file met98.npt is a an ASCII text file that contains the meteorological conditions for the Dead Sea in Israel/Jordan for the year 1998. There are the following data columns:

• JDAY: Julian day (note that Julian day 1.5 is January 1 at 12 noon)
• TAIR: Air temperature (dry bulb) in deg C
• TDEW: dew point temperature in deg C
• WIND: wind speed in m/s
• RAD: wind direction in Radians
• CLOUD: Cloud cover in tenths (from 0-10); 0 no clouds, 10=fully cloudy
• SRO: short wave solar radiation in Watts/m2

This file is used to provide meteorological data to a hydrodynamic and water quality (including temperature) of the Dead Sea. Interesting aspects or questions to ask using the data:

1. What are predominant wind direction and magnitudes for the Dead Sea?
2. What is the frequency of wind events - are they on a semi-diurnal or diurnal or synoptic time scale (every 2 weeks for example)?
3. Can one develop a regression model of air temperatures for the year based on Julian day and solar radiation?
4. What if you averaged out the diurnal period of air temperature, can a reasonable regression be established?
5. Are the various meteorological data correlatable to each other?
   

Transportation Safety Data
Data provided by: Dr. Chris Monsere

The file svror.xls is an Excel workbook file that contains a five year summary of single vehicle run-off- road (SVROR) crashes that did not occur in snow or ice conditions. A run-off-road crash is one where the vehicle departed the travel lane and either crashed into a fixed object (i.e. tree) or overturned. The following fields are included:

• ID: Unique ID field for each record
• ROUTE: Signed highway route (i.e. I-5)
• BMP: Beginning milepoint to section
• EMP: Ending milepoint to section
• TOTAL: Total number of SVROR crashes in 5 year period
• FATA: Number of fatal and severe injury SVROR crashes in 5 year period
• INJ: Number of injury SVROR crashes in 5 year period
• PDO: Number of property damage only SVROR crashes in 5 year period
• ADT: Average daily traffic (average number of vehicles per day on the 5 mile section)
• SHLDR_MID: Average shoulder width
• SHLDR: Width of shoulder (0, 0-3 ft, 4-6ft, >6ft)
• SCS: State classification
• CONDITION: Pavement condition

Interesting aspects or questions to ask using the data:

1. Using all data, can you create a regression model relating volume (ADT) and the total number of crashes (TOTAL)?
2. A common transportation engineering assumption is that increasing shoulder width decreases the number of SVROR crashes. If you add "SHLDR_MID" to your regression model, what are the results? Is the coefficient for the average shoulder width positive or negative?
3. Inspect the relationship SHLDR_MID and ADT. Is there any correlation between ADT and SHLDR_MID? A common problem with safety analysis is that there is a strong correlation between some roadway features and traffic volumes. In this example, high volume roadways typically have wider shoulders. Since the number of crashes is strongly related to the total volume, the effect of shoulder width is difficult to detect in these data
4. Use only data with ADT with volumes between 7000 and 8000. Rerun the model in 2 on these data? Are your results different?
5. Would it better to have actual shoulder widths? Is five miles to long of a section to study?

Environmental Data
Data provided by: Dr. Roy Koch

The file SchollsWQ.xls contains data on water quality and streamflow for the Tualatin River near Scholls, Oregon. The data were collected by Clean Water Services of Washington County. This data set is adjusted; in the original data set occasionally has several observation per day for particular days. This was ofter the result of replicate sampling for quality control. There were also a few periods (principally summers) where there was extensive daily sampling over a short time period. Since the samples are more or less monthly, in every case where there was more than one observartion in a month, they were averaged to produce a single monthly observation. In addtion to the data, the date and time of day of the observation are included.


Intelligent Transportation Systems Data
Data provided by: Dr. Robert Bertini
COMING SOON