Finding #1: Inconsistency in Data

Although the LTPP has an enormous dataset, inconsistency in data collection made accurate analysis and visualization difficult. A total of 2,581 road sections were identified, however the collection of data seemed completely random between year to year, and not every road section is examined every year. In 1996, traffic data was collected on 959 road sections. In comparison, only 160 samples were collected in 2019. In the case of Hawaii, the four identified test sections were sampled a total of 16 times in 20 years, and 1995 was the only year where all four sections were sampled.

Recommendation: Sample the same road sections consistently.






Number of samples collected each year between 1990 and 2019




Finding #2: Road Traffic Distribution


Before diving into the data, we knew Hawaii's footprint would be rather small compared to the national average, given our small land area and population. We were under the assumption states with large populations would have high traffic volume, but to our surprise, we found Georgia, Arkansas, and Oregon consistently ranked the highest in average traffic load. Despite of inconsistency in available data, this finding remains significant.




Finding #3: Time vs. Traffic


As expected, we hypothesized that with the increase of population and the utilization of personal vehicles for work and play, there would be a direct correlation with the amount of traffic on the roads. Through our data analyzation, it is evident that traffic does increase through time. What piqued our interest, is the unusual spike in 2004 that leaves to question the data integrity, or the testimony of an influx of traffic during that year.