This chart comes from LTPP's "TST_L05B" table, which contains data from field and laboratory measurements on material type and thicknesses of the pavement structure layers. The 8 types of classified layers are: Asphalt concrete layer, Bound (treated) base, Bound (treated) subbase, Engineering Fabric, Portland cement concrete layer, Subgrade (untreated), Unbound (granular) base, and Unbound (granular) subbase. For combined layers, the upper-most layer is used. For example, on some AC-surfaced sections with asphalt treated base layers, the combined layer may be represented as AC.

As you can see, asphalt-concrete layer, typically the surface layer, is the most studied layer, and accounts for 39% of all studied road sections, followed by untreated subgrade layer and unbound granular base layer. This can be explained by the likelihood that base layers are rarely accessed during repair.

For Hawaii test sections, only 4 types of layers were collected: Asphalt concrete layer, Bound Portland cement concrete layer, and Subgrade (untreated), Unbound (granular) base.

Please note, a road can have an Asphalt-Concrete surface and Subgrade Base at the same time. Asphalt concrete and Portland Cement Concrete are both surface layers, whereas the latter two are sub-base material.

This chart simply shows that among the chosen road sections participating in LTPP's studies, 41% samples were taken from the asphalt-concrete top layer

We investigated road section crack lengths by state. "Crack Length" is the reinterpreted cracking data for asphalt-concrete surface pavements based on formulas in the Mechanistic-Empirical Pavement Design Guide. This graph shows the average surface layer crack length for each state. Based on analysis, we found it interesting that the Midwest States -- North Dakota, Minnesota, South Dakota, Indiana, and Wyoming-- seem to have the highest average crack length. Intuitively, this makes sense as those states experience harsh winters.

Page 55-57 of the LTPP User Guide explains exactly how crack length is measured, calculated, and interpreted.

The User Guide does not define a unit for crack length. However, crack length relativity remains the same, a crack length of 10 is a bigger crack than a crack length of 5, by approximately two times.

Next, we attempted to examine the correlation between a road section's representation thickness and its crack length. After having performed the analysis above, we realized that we cannot draw conclusions on representative thickness and crack length.

Representative thickness is a measure of the test section's thickness at a specific layer, this could be the surface layer, the base, or the sub-base. Whereas the the crack length is only measured at the topmost layer of pavement.

Further, the samples of thickness and samples and crack length were collected at different times. If a road section's sub base layer is measured at 20.5 mm in 2005, and surface crack length is measured to be 16" in 2007, we cannot draw any conclusions between the two sets of data.

This is another example of LTPP's large collection of inconsistent, and uncorrelated data.

Equivalent single-axle load (ESAL) is the unit used to measure the effects of axle loads on pavement.

By convention, an 18,000-pound single axle is 1.00 ESAL. 1 KESAL is 1,000 ESAL. The KESAL chart comes from LTPP's TRF_ESAL_COMPUTED table, which captures the estimated axle load a lane section experiences in the given year based on traffic monitoring measurements computed using the 1993 AASHTO Guide for Design of Pavement Structures methodology. Data is available for all 50 states, ranging from 1990 to 2019, although not every State collects data every year.

Our analysis shows, on average, states with the heaviest traffic are: Georgia, Arkansas, and Oregon, which is a bit surprising because none of these states has a particularly large population. Upon a closer look, we realized that only 6 years of data are available for Georgia. Since the middle of the year range has more data, we took a closer look at the year 2000, which shows the states with the most traffic are: Indiana, Oregon, and Pennsylvania. Oregon was again among the top 3 in both sample sets, which begs a closer analysis.

More information on the measure of KESAL can be found here

We decided to examine the correlation between time and traffic load for the state of California, since this state has the most data and roads are sampled most consistently.

Despite the inconsistency in data, we do see an upward trend in road traffic and time progresses.

Hawaii Traffic Load by Road ID

We wanted to see what traffic load for Hawaii looked like. During data processing, we realized that there are only 4 test sections in the state of Hawaii: 2 on Big Island, and 2 on Maui. The two test sections on Maui are so closely located they are practically the same section. Further, not every road section is sampled every year. That said, the sample size of test sections in Hawaii is too small to make definitive conclusions.

Hawaii Average Annual Traffic Load

We examined traffic loads in Hawaii over time. We were hoping to see an increasing trend in road traffic in Hawaii, however, since the sample size is so small, we do not have sufficient data to make such a conclusion.