CHAPTER 4

Conclusions and Suggested Research

4.1 Conclusions

Environmental conditions can have a significant effect on flexible and rigid pavement performance. Factors such as temperature, precipitation, cloud cover, and freezing index are important in identifying the impact of the environment on a pavement section. These factors not only affect how the pavement layer materials behave when subjected to environmental loadings, but they also affect pavement layer responses to traffic loadings. The climate-related inputs and algorithms included in the AASHTOWare Pavement ME Design’s climate model (i.e., EICM) were reviewed to identify and address potential limitations. The MERRA-2 assimilated dataset includes a large number of climate-related variables that could be used to help improve the EICM prediction capabilities. Several enhancements were developed to address the limitations. The proposed enhancements and future research topics are summarized in the next section.

4.2 Suggested Research

4.2.1 Summary of Enhancements and Implementation Steps

4.2.1.1 Additional MERRA-2 Variables

Based on this research, the following additional variables from the MERRA-2 database related to the overall energy balance are suggested:

• High-priority variables: Hourly shortwave radiation, longwave radiation variables from the MERRA-2 LAND Diagnostics Table
• Lower-priority variables: Hourly latent heat, sensible heat, and ground heat variables from the MERRA-2 LAND Diagnostics Table

Currently, the EICM/PMED selects and downloads the hourly climate data directly from the LTPP InfoPave website. The data obtained from LTPP InfoPave is formatted to what the PMED expects. The potential steps to make the additional variables available for direct user consumption are summarized here:

1. FHWA or LTPP: Add additional variables to the LTPP InfoPave Climate Tool.
2. FHWA or LTPP, in collaboration with AASHTO/AASHTOWare: Update or revise the HCD file format to include the hourly shortwave and longwave radiation variables in addition to the currently available data (e.g., air temperature, wind speed, relative humidity, percent sunshine, and precipitation).
3. AASHTOWare: Update the parsing logic to accommodate the additional hourly variables in the PMED software.

4. AASHTOWare: Update the EICM source code to use the MERRA-2 shortwave and longwave radiation data to calculate the total net radiation, which replaces the empirical equation. The results will affect the predicted pavement temperatures, which will likely require a recalibration of the performance prediction models for rigid and flexible pavements.
5. AASHTOWare: Implement the updated EICM after global calibration is completed.

4.2.1.2 Other Proposed Changes for Consideration

• Improve documentation of the models.
• Improve documentation of inputs and outputs.
  • – Input file improvements
    • Standardize the input format and structure.
  • – Output file improvements
    • Standardize the output file format of the intermediate output files.
    • Provide an option to export additional calculated values for each node. Examples of such variables include the predicted pavement temperature, moisture content, pore pressure, resilient modulus adjustment factors, frost depth, and thaw depth.
    • Provide an option to specify the time increments for the output data. Examples include hourly, daily, monthly, or yearly. These additional time increments can be useful to other applications beyond the EICM/PMED, especially within applications that focus on resiliency and sustainability.
• Make the EICM a standalone application or web service to foster its usage beyond the AASHTOWare PMED. One example is to allow other applications to run the EICM and use the results. A standalone application can also help with future research studies specifically targeted to be implemented into the PMED software. Modern software code repository tools can help facilitate these types of integration efforts.
• Allow integration and collaboration between other entities. This consideration builds upon the collaboration between AASHTOWare and LTPP InfoPave and expands it.

4.2.2 Proposed Changes to the AASHTO MEPDG MOP

The AASHTO MEPDG MOP does not currently provide an in-depth overview of all the equations, models, and methods included in the AASHTOWare PMED software and focuses more on the general design method and procedures that apply to mechanistic empirical pavement design.

The following items summarize the anticipated changes or additions to the MOP:

• Expansion of climate section to include the results from this research.
• Addition of an appendix or supplement document highlighting the EICM equations and methods.

4.2.3 Future Research Recommendations

The amount of climate-related research and available data is developing rapidly. The following research activities are recommended:

• Investigate the use of different climate data sources other than MERRA-2.
  • – Other global assimilated datasets include Copernicus.
• Investigate a procedure to use forecasted climate data for pavement design.
  • – This procedure can be based on a time series model analysis, such as autoregressive integrated moving average or other statistical methods.

• – Use the historical climate data and select a subset of values based on the 50th, 75th, or 90th percentile values. The impact on the predicted pavement performance can help quantify how resilient the pavement design is compared to extreme cases.
• – The challenge with forecasted data is the time it takes to generate hourly data for an entire pavement design period. Will it be generated ahead of time and stored in a database for user access, or will it be performed in real time when it is requested or imported into a design software?
• – This procedure could also be a collaborative effort between multiple government agencies, such as NASA, FHWA, AASHTO, and others who may find a need for long-term forecasted data.
• Research methods to create “custom climate” data to artificially introduce extreme events into the climate data used for analysis.
  • – For example, adjust the climate data to include a number of randomly spaced extreme events over a design or analysis period. One event could consist of a prolonged extreme rainfall event over a 7-day period; another event could consist of extreme heat spanning a month or more or an extreme cold temperature event in a region where it is not typically expected, similar to The Great Texas Freeze in 2021. Such a feature would make it possible to study the impact of these events on the predicted pavement performance from a resiliency standpoint.
• Determine a new model or regression equation that provides an indication of the “deep ground temperature” to improve the initial boundary conditions of the EICM.
• For NCHRP projects that produce a software application, or any type of source code, analysis code, and so forth, require the use of a source code version control system and code repository. Version control systems are essential to tracking code changes, working collaboratively with other researchers or users, and setting up test cases and expected behaviors. Popular code repository systems include GitHub, Bitbucket, GitLab, and Azure DevOps. Many of these systems can be used to create both private and open-source code repositories.
• Further investigate interpolation methods to create a database of “climate locations” along major roadways or roadway networks, as shown in Figure 37. The red point locations represent the available MERRA-2 gridpoint locations, the blue point locations represent the proposed climate location points that follow along known roadways and/or interstates, and the single yellow point location is where the pavement design is performed or the “project” is located. The gridded MERRA-2 locations or any other gridded assimilated dataset can be used to interpolate specific points linked to a highway network.