Several traffic-related inputs other than VMT by time-of-day and average speed are needed by MOBILE5a. We are formulating models and methods to improve the estimation of these inputs. For example, we have developed a fractional split that determines the VMT mix ratio as a function of several, informative, variables.  Thus, rather than using only a binary distinction of freeways or arterials, we are also using more details of the geometrics of the roadway segment such as number of lanes, width of lanes, . Similarly, instead of a simple binary distinction between rural vs. urban area type, we are using more disaggregate and continuous indicators of the degree of urbanization of each traffic survey zone (TSZ). Such refinements will allow more accurate VMT mix ratio computations and at a finer level of geographic resolution. These ratios can be applied to each link in the network based on its attributes and on its geographic location to obtain link-specific VMT by time-of-day by vehicle type. Other traffic-related inputs to MOBILE5a for which we are formulating models include vehicle age and accumulation rate and VMT on local roads.