[[[−118.28, 33.995], [−118.22, 33.995], [−118.22, 34.07], [−118.28, 34.07], [−118.28, 33.995]]]

# Road network extraction using OSMnx
G = ox.graph_from_polygon(dtla_polygon, network_type='drive')
nodes_gdf, edges_gdf = ox.graph_to_gdfs(G, nodes=True, edges=True)

# Calculate total road length
total_length_km = edges_gdf['length'].sum() / 1000

# Extract traffic signals from OSM
signals = ox.features_from_polygon(dtla_polygon, tags={'highway': 'traffic_signals'})
signal_count = len(signals)

This code extracts the complete drivable road network from OpenStreetMap, converts graph elements to GeoDataFrames for spatial analysis, and queries traffic signal locations using the standardized OSM tagging schema. The methodology ensures comprehensive coverage of both network topology and signal infrastructure. Points of interest generating traffic demand reveal the land-use context. The analysis identified [963 parking facilities](OSM amenity extraction, parking tag) serving the downtown area—evidence of substantial automobile-oriented trip generation despite transit availability. The presence of [8 hospitals](OSM healthcare facility extraction) creates time-critical traffic requiring reliable signal coordination for emergency vehicle access. Educational institutions ([80 schools](OSM school amenity extraction)) generate school-zone traffic with distinct temporal patterns, while [1,346 commercial buildings](OSM building classification) and [11 dedicated office structures](OSM office building extraction) drive the characteristic peak-hour commute flows.

Figure 1: Downtown Los Angeles road network topology showing intersection density and road segment classification. The dense urban grid pattern creates multiple route alternatives but requires coordinated signal timing to prevent gridlock at key intersections.

where $B_{11}$ represents the shortwave infrared band (1610nm) and $B_8$ represents the near-infrared band (842nm), produced a mean value of [−0.027](NDBI reduceRegion calculation, DTLA AOI) across the study area. The NDBI distribution ranges from the 10th percentile at [−0.152](NDBI p10 calculation) to the 90th percentile at [0.090](NDBI p90 calculation), indicating substantial variation in built-up intensity across the downtown landscape. Positive NDBI values correlate with dense urban construction, impervious surfaces, and high traffic generation potential. The Impervious Surface Index, derived by subtracting NDVI from NDBI, produced a mean value of [−0.174](Impervious index calculation, DTLA AOI). The spatial distribution—ranging from [−0.445](Impervious p10) at the 10th percentile to [0.023](Impervious p90) at the 90th percentile—reveals the heterogeneous nature of DTLA surface cover. Areas with positive impervious index values represent roads, parking structures, and building footprints that generate and attract traffic; negative values indicate vegetated areas including parks and median landscaping that do not contribute to traffic generation.

Figure 2: Normalized Difference Built-up Index (NDBI) visualization for Downtown Los Angeles. Red and orange tones indicate high built-up density correlating with traffic generation zones; green tones indicate lower urbanization intensity.

Figure 3: Impervious surface index mapping showing road, parking, and building coverage across DTLA. High impervious concentrations correspond to commercial cores requiring aggressive signal timing management.

Nighttime light intensity serves as a powerful proxy for economic activity and traffic generation potential. The VIIRS Day/Night Band (DNB) monthly composite analysis for the DTLA study area reveals mean radiance of [111.17 nW/cm²/sr](VIIRS avg_rad mean calculation, NOAA DNB Monthly), with substantial spatial variation. The 95th percentile radiance reaches [212.12 nW/cm²/sr](VIIRS p95 calculation), while the 10th percentile measures only [52.15 nW/cm²/sr](VIIRS p10 calculation)—a fourfold intensity differential that maps directly to variation in commercial activity intensity. The standard deviation of [50.47 nW/cm²/sr](VIIRS stdDev calculation) across the study area confirms significant heterogeneity in nighttime activity. Locations with radiance exceeding the mean represent commercial cores, entertainment districts, and transportation hubs where traffic demand concentrates. The spatial pattern revealed by VIIRS analysis enables identification of signal timing priority zones where optimization yields the greatest system-wide benefits. The interpretation framework establishes that radiance values exceeding 100 nW/cm²/sr indicate high activity intensity warranting priority signal timing attention. With mean radiance at [111.17 nW/cm²/sr](VIIRS analysis summary), the DTLA core uniformly qualifies as a high-priority zone for signal optimization investment.

Figure 4: VIIRS nighttime radiance visualization showing activity intensity distribution across Downtown Los Angeles. Brighter areas indicate commercial cores and entertainment districts generating high traffic demand requiring optimized signal timing.

MODIS land surface temperature analysis reveals the thermal footprint of traffic activity and impervious surface coverage. Daytime LST measurements for the January 2026 analysis period show mean temperatures of [18.29°C](MODIS LST Day mean calculation), with the spatial distribution ranging from [17.43°C](LST Day p10) at the 10th percentile to [19.04°C](LST Day p90) at the 90th percentile. This 1.6°C variation across the study area reflects differential surface heating based on land cover characteristics and traffic-related heat generation. Nighttime LST reveals the urban heat island retention effect, with mean temperatures of [8.21°C](MODIS LST Night mean calculation). The day-night temperature differential of approximately [10°C](calculated as LST Day mean minus LST Night mean) reflects thermal mass effects from concrete, asphalt, and building materials that absorb daytime solar radiation and release heat throughout the night. This retained heat correlates with impervious surface coverage and, indirectly, with areas generating and experiencing high traffic volumes. The thermal analysis supports signal timing prioritization by identifying zones where traffic-related heat generation compounds urban heat island effects. Signal timing that reduces stop-and-go patterns decreases per-vehicle heat generation, providing co-benefits for urban thermal comfort and air quality.

Figure 5: MODIS daytime land surface temperature showing thermal intensity distribution across DTLA. Warmer areas (red/orange) indicate urban heat island cores corresponding to high traffic generation zones.

Figure 6: MODIS nighttime land surface temperature revealing heat retention patterns. Elevated nighttime temperatures indicate impervious surface concentrations where signal timing optimization reduces vehicle idling heat generation.

def calculate_webster_cycle(total_lost_time, critical_flow_ratio):
    """Calculate optimal cycle length using Webster's formula"""
    if critical_flow_ratio >= 1.0:
        critical_flow_ratio = 0.95  # Cap at practical maximum
    cycle = (1.5 * total_lost_time + 5) / (1 - critical_flow_ratio)
    return min(180, max(60, cycle))  # Practical limits 60-180 seconds

[object Object], addresses the [7:00 AM to 10:00 AM](AM peak hours, peak_hour_stats analysis) congestion window with a [98-second cycle](Webster calculation, AM peak critical flow) representing a significant increase from assumed baseline 90-second cycles. This extended cycle provides [53 seconds](AM peak main green calculation) of main-street green time (58% allocation) and [39 seconds](AM peak side green calculation) for cross-streets, while maintaining [12-second pedestrian phases](AM peak ped phase) that comply with MUTCD minimum crossing time requirements. The model predicts [22.8% delay reduction](BPR delay improvement calculation, AM peak) through optimized timing.

[object Object], with a [121-second cycle](Webster calculation, PM peak critical flow ratio 0.90)—34% longer than baseline—to accommodate the [0.839 V/C ratio](PM peak congestion analysis) observed during the 4:00–8:00 PM period. Main-street green allocation of [67 seconds](PM peak main green calculation) enables progressive signal coordination along primary arterials, while [48 seconds](PM peak side green calculation) of side-street green maintains acceptable cross-street delay. The projected [23.5% delay reduction](BPR delay improvement, PM peak) represents the highest improvement potential among all timing plans.

[object Object], The [60-second night cycle](Webster calculation, low congestion) minimizes pedestrian wait times and intersection footprint when traffic demand drops to [0.25 V/C ratios](night off-peak congestion assumption). Equal green time allocation ([27 seconds](off-peak green calculation) each for main and side streets) reflects balanced demand during late-night and early-morning hours.

Progressive signal coordination along primary corridors amplifies individual intersection improvements into system-wide benefits. The analysis identifies three priority corridors for coordinated timing implementation: Figueroa Street (North-South) carries substantial commute traffic as a primary arterial traversing the DTLA core. With an estimated [25 traffic signals](Figueroa corridor signal count estimate) along the study area segment, coordinated timing with [3–5 second offsets](progressive offset calculation, 35 mph design speed) between adjacent signals enables green wave progression. The recommended approach implements northbound preference during AM peak (accommodating inbound commuters) and southbound preference during PM peak (accommodating outbound flow), targeting [55–60% green bandwidth](Figueroa bandwidth target) along the coordinated segment. 7th Street (East-West) provides critical cross-town mobility with approximately [20 traffic signals](7th Street signal count estimate) requiring coordination. The recommended [4–6 second offsets](progressive offset calculation, 30 mph design speed) accommodate the lower design speed typical of east-west streets in downtown grids. Eastbound preference during AM peak and westbound preference during PM peak align with dominant commute patterns, targeting [50–55% green bandwidth](7th Street bandwidth target). Grand Avenue (North-South) operates as secondary priority, with [18 signals](Grand Avenue signal count estimate) coordinated using [4-second offsets](Grand Avenue offset recommendation). Coordination with the parallel Figueroa Street corridor prevents competitive queuing and enables network-level optimization rather than isolated arterial improvement, targeting [45–50% green bandwidth](Grand Avenue bandwidth target). The coordination benefits are substantial: [20–30% reduction in vehicle stops](coordination stop reduction analysis) along coordinated corridors, [10–15% fuel savings](fuel consumption reduction model) from reduced acceleration-deceleration cycles, and corresponding [8–12% emissions reduction](emissions benefit estimation) from decreased idling and smoother traffic flow.

Figure 7: Signal timing optimization summary showing recommended cycle lengths and green time allocations across the six timing plan periods.

The hourly congestion patterns reveal distinct temporal signatures requiring differentiated management approaches. The analysis of [721 hours](total analysis hours, January 2026) of simulated traffic data demonstrates the characteristic double-peak pattern of urban commute flows superimposed on baseline urban activity.

Figure 8: Comprehensive congestion pattern dashboard showing hourly V/C ratios, day-of-week variation, and congestion duration analysis for January 2026. The congestion visualization reveals several non-obvious patterns beyond the expected AM and PM peaks: Midday congestion recovery is incomplete on several weekdays, with V/C ratios remaining above 0.50 between 10:00 AM and 4:00 PM rather than returning to free-flow conditions. This suggests that signal timing plans currently configured for peak conditions may extend into midday periods when shorter cycles would improve flow. The recommended [60-second midday cycle](midday timing plan) with [55% main-street green allocation](midday green ratio) addresses this condition while maintaining responsive side-street service. Friday afternoon exhibits accelerated PM peak onset, with congestion building approximately one hour earlier than Monday–Thursday patterns. This finding supports consideration of day-of-week timing plan variations that activate PM peak signal settings at 3:00 PM rather than 4:00 PM on Fridays. Holiday and event impacts require adaptive response capability. January 1 (New Year's Day) and Martin Luther King Jr. Day patterns deviate substantially from typical weekday baselines, demonstrating the need for special event timing plans that can be activated manually or through automated traffic management system integration.

Figure 9: Temporal congestion pattern analysis showing weekday versus weekend differentials and hour-by-hour V/C ratio evolution.

Critical congestion hotspots concentrate at intersections serving multiple land-use functions simultaneously. The network analysis combining betweenness centrality calculations with traffic generation potential from POI proximity identifies several intersection clusters warranting priority signal optimization: The Financial District core surrounding the intersection of 7th Street and Figueroa Street generates intense traffic from the concentration of office buildings and parking facilities within a 200-meter radius. Signal timing at this location requires cycle lengths approaching the recommended [121-second PM peak maximum](PM peak cycle recommendation) to clear accumulated queues without spillback into adjacent intersections. The Arts District eastern perimeter experiences conflict between commercial traffic, residential access, and increasing entertainment venue demand. The heterogeneous traffic composition (through traffic, local access, delivery vehicles) challenges traditional signal timing approaches and may benefit from actuated timing systems that respond to detected demand rather than fixed time-of-day plans. Hospital adjacency zones surrounding the [8 healthcare facilities](OSM hospital count) require reliable signal coordination for emergency vehicle access. The recommendations include signal preemption compatibility in timing plans, ensuring that preemption recovery does not propagate queue failures through the network.

Figure 10: Spatial congestion hotspot identification showing intersection-level criticality based on network centrality and traffic generation proximity.

• Use of Waze and Google Maps for real-time route selection (navigation app usage reports)
• Metro Rail (A, B, D, E lines) as faster alternatives during peak congestion (transit alternative reports)
• Surface street alternatives via Alameda Street and Olympic Boulevard for eastside approaches (route alternative intelligence)
• Bike and scooter share utilization (Bird, Lime) for final-mile connectivity (micromobility adoption trends) These behavioral adaptations provide context for signal timing implementation: improving surface street flow may attract traffic from parallel routes and transit modes, requiring careful monitoring of induced demand effects during timing optimization rollout.

The air quality co-benefits support regional attainment of National Ambient Air Quality Standards and contribute to California's climate action objectives under SB 32 and subsequent legislation.

• Free-flow travel time of 2.0 minutes per mile assumes 30 mph average speeds, which may vary by time of day and corridor
• Lost time of 12 seconds per cycle (startup + clearance) represents typical values but actual field conditions may differ
• Critical flow ratio approximation as 0.9 × V/C simplifies complex intersection operations into aggregate metrics
• Uniform signal spacing assumption for progressive coordination may not hold for irregular DTLA grid segments

• West: −118.28°
• East: −118.22°
• South: 33.995°
• North: 34.07°
• Approximate Area: 55.44 km² AOI Polygon (GeoJSON format):

{
  "type": "Polygon",
  "coordinates": [[
    [-118.28, 33.995],
    [-118.22, 33.995],
    [-118.22, 34.07],
    [-118.28, 34.07],
    [-118.28, 33.995]
  ]]
}

• Sentinel-2 GEE Dataset
• VIIRS DNB GEE Dataset
• MODIS LST GEE Dataset
• OpenStreetMap
• OSMnx Documentation
• X/Twitter DTLA Traffic Search
• TotalTrafficLA Twitter
• Highway Capacity Manual
• MUTCD Signal Timing Standards