It is possible to develop a one-dimensional thermal model that can forecast with some precision when
vehicular load restrictions should initially be implemented as well as removed. The model forecasts
could then be a complement to the current RWIS TDP data that are used to determine the timing of load
restrictions. In order to produce forecasts quickly and unambiguously such that it can be easily
integrated into the current RWIS online data dissemination system, the model would be based on a
reduced set of important physical properties and phenomena and implemented in one dimension only.
The proposed numerical model would be a one-dimensional finite difference thermal energy balance
that accounts for both sensible and latent heat effects in a saturated/semi-saturated soil. The ultimate
goal is to accurately forecast then time when specific vertical locations thaw; specifically 1 foot and 5
feet below surface in this case.
Initial input parameters into the model would include the physical properties of the underlying soil
materials. Effective freezing and thawing indices (n-factors) will be determined using archived RWIS
data of temperature profiles over time. Effective surface temperatures can then be matched to air
temperatures for a top-surface boundary condition into the one-dimensional model.
Initial validation of the model would involve comparison of the model forecasts with several years of
archived Alaska RWIS data from TDP measurement sites. We will work with Alaska DOT&PF for
acquisition of the data, and to gain an understanding of the conditions at each of the RWIS locations.
The model validation metrics would be deviation between forecast and recorded temperatures with
focus on temperatures near freeze/thaw. Model design iteration would include reducing the model input
requirements such that it can be modified and implemented efficiently for use in a wide range of
locations. The ulterior motive of this step is to be applicable to locations far from any current RWIS
installations. After the development, validation, and iteration steps, the forecast system would be
designed and packaged for facile integration with the current Alaska RWIS data system.