Chore 🧹: Evaluate Highway Skim Segmentation for DA, SR2, and SR3+ Modes

[i-am-sijia]

Background
TDM23 currently provides the following highway skims:

• {da/sr}__toll__{AM/MD/PM/EV}
• dist__{AM/MD/PM/EV}
• {da/sr}__time__{AM/MD/PM/EV}
• rs_fare__{AM/MD/PM/EV}

The congested toll and travel time skims are segmented by Drive Alone (DA) and Shared Ride (SR). However, the congested distance skim is not segmented by auto mode.

Since the mode choice model includes three auto modes:

• DA
• SR2
• SR3+

Ideally, the highway skims should support the same level of segmentation as the mode choice model. While a single shared-ride skim may be sufficient for the base year, future scenarios may include facilities, toll policies, or pricing structures that differentiate between SR2 and SR3+ vehicles. The current skim structure would not support such distinctions.

Additional skim segmentation will increase memory requirements and increase Lighthouse runtime and resource usage.

Goals

1. Evaluate the appropriate segmentation scheme for highway skims based on current and future modeling needs.
2. Determine whether separate skims should be generated for: DA, SR2, and SR3+
3. Assess the computational and memory impacts of additional skim segmentation.
4. Generate and integrate the revised highway skims if warranted.