Estimating Traffic Volume Growth: Practical Methods and Forecasting Techniques

by

Table of Contents

Estimating traffic volume growth is a critical component of modern transportation planning, infrastructure development, and urban management. As cities expand and populations increase, understanding how traffic patterns will evolve becomes essential for making informed decisions that shape our transportation networks. Accurate traffic forecasts enable planners, engineers, and policymakers to anticipate future demand, allocate resources efficiently, and design infrastructure that meets the needs of growing communities while maintaining safety and operational efficiency.

Through the traffic forecasting process, planners, engineers, and other transportation professionals must estimate the amount of traffic that will exist on the transportation system in the future and assess the short- and long-term impacts that traffic will have on the transportation network. This complex undertaking requires a combination of analytical methods, data collection techniques, and forecasting models that can account for numerous variables affecting traffic patterns. The stakes are high—getting traffic forecasts right is crucial to make critical project decisions about structural and geometric capacity, financial viability, and environmental effects.

Understanding Traffic Volume Metrics

Before diving into forecasting methods, it’s essential to understand the fundamental metrics used in traffic analysis. The most commonly used measure is Annual Average Daily Traffic (AADT), which serves as the foundation for most traffic growth estimates and planning decisions.

Annual Average Daily Traffic (AADT)

Annual average daily traffic (AADT) is a measure used primarily in transportation planning, transportation engineering and retail location selection. Traditionally, it is the total volume of vehicle traffic of a highway or road for a year divided by 365 days. This metric provides a standardized way to compare traffic volumes across different locations and time periods.

Annual Average Daily Traffic (AADT) represents the average daily traffic volume across an entire year. It accounts for weekdays, weekends, holidays, and seasonal fluctuations, providing a standardized number. This comprehensive approach makes AADT particularly valuable for long-term planning and infrastructure investment decisions.

AADT is the standard measurement for vehicle traffic load on a section of road, and the basis for some decisions regarding transport planning, or the environmental hazards of pollution related to road transport. One of the most important uses of AADT is for determining funding for the maintenance and improvement of highways. In the United States, federal funding allocations to states are directly tied to traffic volumes measured across highway networks, making accurate AADT calculations financially significant.

Average Daily Traffic (ADT) vs. AADT

While AADT and ADT are often confused, they represent different measurement periods. AADT is the total volume of vehicle travel on a road for an entire year, divided by 365. ADT is the average number of vehicles traveling through a location during a period shorter than a year. ADT might be calculated for a specific season, month, or week, making it useful for understanding short-term traffic patterns or seasonal variations.

AADT is generally used to measure long-term trends or changes in travel demand, while ADT is more useful for short-term planning and operations. Both metrics serve important but distinct purposes in transportation planning and analysis.

Data Collection Methods for AADT

There are two different techniques of measuring the AADTs for road segments with automated traffic counters. One technique is called continuous count data collection method. This method includes sensors that are permanently embedded into a road and traffic data is measured for the entire 365 days. This approach provides the most accurate AADT measurements but requires significant investment in equipment and maintenance.

Most AADTs are generated using short-term data collection methods sometimes known as the coverage count data collection method. Traffic is collected with portable sensors that are attached to the road and record traffic data typically for 2 – 14 days. These short-term counts are then adjusted using expansion factors derived from permanent count stations to estimate annual averages.

Fundamental Methods for Estimating Traffic Growth

There are various analysis methods that depend on factors such as the complexity and size of the project, the surrounding environment and access to the project, and prospective density and growth of the study area. Planners and engineers apply forecasting techniques to projects of all sizes and complexities, ranging from single intersections to statewide recommendations. The selection of an appropriate forecasting method depends on data availability, project scope, required accuracy, and the planning horizon.

Historical Data Analysis and Trend Projection

Analyzing historical traffic data provides the foundation for most traffic growth estimates. This method involves examining past traffic counts to identify patterns and trends that can be projected into the future. For forecasting without TDM outputs, growth patterns and historical trends are used as a means of estimating future volumes.

Reliable growth rate selection needs solid historical traffic data as its foundation. Most transportation agencies prefer to use traffic count information that is 20+ years old to create reliable trend lines. Long-term historical data helps smooth out short-term fluctuations and provides a more stable basis for projecting future trends.

When analyzing historical trends, transportation professionals must account for economic cycles and unusual events. Economic fluctuations need special attention during trend analysis. Traffic growth often shows temporary dips during recessions or downturns that could skew projections. Long-term analysis shows minimal effects from short-term economic downturns, though extended slow-growth periods need careful thought.

Linear Growth Rate Formula

The linear growth rate formula is one of the simplest and most commonly used methods for traffic forecasting. Linear Growth – Linear increase in traffic volumes over time. This method assumes a constant amount of growth in each year and does not consider a capacity restraint. This approach works well for short-term forecasts and areas with stable, predictable growth patterns.

The basic linear formula calculates future traffic volume by applying a constant growth rate to the base year volume. For example, if a roadway currently carries 24,000 vehicles per day with an expected 2% annual growth rate, the projected volume after five years would be calculated as: Future Volume = Base Volume × (1 + Growth Rate × Years). This would result in 26,400 vehicles per day, representing a total increase of 2,400 vehicles over the five-year period.

Federally funded projects and environmental reviews typically require the projection of traffic volumes 10–30 years in the future, typically assuming a 1–2% annual growth in vehicle volume. These standard growth rates have been widely adopted across the transportation planning industry, though they may not always reflect actual conditions in specific locations.

Compound Growth Rate Method

Unlike linear growth, the compound growth method assumes traffic increases exponentially over time, with growth building upon previous years’ increases. A 2% compound traffic growth rate doubles traffic in 35 years. This dramatic long-term effect demonstrates why the choice between linear and compound growth methods can significantly impact infrastructure planning decisions.

The compound growth formula is: Future Volume = Base Volume × (1 + Growth Rate)^Years. This exponential calculation can produce substantially different results than linear growth over longer time periods, making it important to select the appropriate method based on local conditions and historical patterns.

The accuracy of future projections and infrastructure planning decisions depends heavily on choosing the right growth rates between linear and compound formulas. Transportation planners must make this crucial decision while forecasting traffic volumes for 2025-2030.

Declining Growth Curves

In areas approaching capacity constraints, simple linear or compound growth models may overestimate future traffic volumes. Areas that have or will likely have capacity constraints should use the declining growth curve shown below as long as there is sufficient evidence of a potential change. Declining growth curves recognize that as roadways approach their capacity limits, traffic growth naturally slows.

In many cases a linear growth rate is used since often there is insufficient data to support use of a more specific type of curve. However, when data supports it, declining growth curves provide more realistic forecasts for congested urban areas.

The growth rate for the first 20 years is typically higher than the growth rate for more than 20 years because the growth will taper down over time as the facility approaches capacity. This recognition of capacity constraints is essential for accurate long-term forecasting.

Trend-Line Forecasting

A trend-line forecast might be made, assuming that the recent percentage of growth in traffic will continue in the future. These trend-line forecasts are most reliable for relatively short periods of time (5 years or less). This straightforward approach extrapolates recent trends forward, making it accessible and easy to implement.

However, trend-line forecasting has important limitations. They do not take into account the potential of future capacity constraints to restrict the growth of future demand. This means trend-line forecasts may overestimate future traffic in areas where roadway capacity will limit growth.

Advanced Forecasting Techniques

While simple historical trend analysis provides a baseline for traffic forecasting, more sophisticated methods incorporate additional variables and modeling techniques to improve accuracy and account for complex interactions between transportation systems and land use patterns.

Travel Demand Modeling

Forecasts of future travel demand are best obtained from a travel demand model. These models require a great deal of effort and time to develop and calibrate. If one does not already exist, then the analyst may seek to develop demand forecasts based on historic growth rates. Travel demand models represent the most comprehensive approach to traffic forecasting, though they require substantial resources to develop and maintain.

Travel demand modeling includes the selection of an applicable model, calibration to local conditions, validation of model results, and revisions of forecast volumes. A TDM output is calibrated for regional trip generation, trip distribution, mode choice, and assignment. This four-step process forms the foundation of most regional transportation planning efforts.

The four-step travel demand modeling process includes:

  • Trip Generation: Estimating the number of trips produced by and attracted to different zones based on land use characteristics, population, employment, and other socioeconomic factors
  • Trip Distribution: Determining where trips that originate in one zone will end, creating origin-destination patterns
  • Mode Choice: Predicting which transportation mode (car, transit, bicycle, walking) travelers will use for their trips
  • Trip Assignment: Assigning trips to specific routes through the transportation network based on travel times, costs, and other factors

TDM outputs can have inconsistencies on a link-by-link basis since these models are not calibrated at the link level. Therefore, further adjustments or post-processing of the model’s daily or peak-period outputs are typically applied prior to use in corridor or project analysis. This post-processing helps reconcile model outputs with observed traffic counts and local conditions.

For projects within the boundaries of an MPO, the latest MPO TDM outputs can be referenced to determine corridor growth rates for projects. For projects outside MPO boundaries, TPP’s SAM can be referenced to determine growth rates. This hierarchical approach ensures that forecasts leverage the best available modeling resources for each location.

Time Series Analysis and Statistical Methods

Time series analysis applies statistical techniques to historical traffic data to identify patterns and forecast future values. These methods can capture seasonal variations, trends, and cyclical patterns that simpler approaches might miss. Common time series techniques include moving averages, exponential smoothing, and autoregressive integrated moving average (ARIMA) models.

ARIMA models are particularly useful for traffic forecasting because they can account for both trends and seasonal patterns in the data. These models analyze the autocorrelation structure of traffic data—how traffic volumes at one time period relate to volumes at previous time periods—to generate forecasts.

Regression Analysis

Regression analysis establishes mathematical relationships between traffic volumes and explanatory variables such as population, employment, income levels, fuel prices, and land use characteristics. By quantifying these relationships using historical data, regression models can forecast future traffic based on projections of the explanatory variables.

Multiple regression models can incorporate numerous variables simultaneously, allowing analysts to account for the complex factors that influence traffic growth. For example, a regression model might predict traffic volumes based on population density, employment in the corridor, household income, distance to the central business district, and the availability of transit alternatives.

It is recommended that TDM outputs and growth rates be compared with other sources, such as historical data, adjacent projects, past projects, state demographer forecasts, and US Census data population. This validation against multiple data sources helps ensure forecast reliability.

Machine Learning and Artificial Intelligence Approaches

Advancements in Machine Learning (ML) and Artificial Intelligence (AI), as well as improvements in Internet of Things (IoT) sensor technologies have made TCP research crucial to the development of Intelligent Transportation Systems (ITSs). These emerging technologies are transforming traffic forecasting capabilities.

While statistical models handle simple patterns well, machine learning, deep learning and ensemble techniques are better suited for complex and changing traffic conditions, though they come with challenges like overfitting and high computational demands. Machine learning models can identify non-linear relationships and complex patterns that traditional statistical methods might miss.

Neural networks, support vector machines, and random forest models have all been applied to traffic forecasting with promising results. These models can process large volumes of data from multiple sources and adapt to changing conditions over time. However, they require substantial training data and computational resources, and their “black box” nature can make it difficult to understand and explain their predictions.

Simulation and Microsimulation Techniques

Traffic simulation models create detailed representations of traffic flow, allowing analysts to test different scenarios and evaluate the impacts of proposed changes. Microsimulation models simulate individual vehicle movements through the network, accounting for driver behavior, vehicle characteristics, and traffic control devices.

These models are particularly useful for analyzing complex intersections, evaluating operational improvements, and assessing the impacts of new developments. While microsimulation requires detailed input data and significant computational resources, it provides insights into traffic operations that aggregate models cannot capture.

Key Factors Influencing Traffic Volume Growth

Accurate traffic forecasting requires understanding the diverse factors that drive changes in travel demand. These factors operate at multiple scales—from individual travel decisions to regional economic trends—and interact in complex ways.

Population Growth and Demographics

Population growth is one of the most fundamental drivers of traffic volume increases. As more people live in an area, more trips are generated for work, shopping, school, and other activities. However, the relationship between population and traffic is not always linear—demographic characteristics matter significantly.

Age distribution affects travel patterns, with working-age adults typically generating more vehicle trips than children or retirees. Household size influences trip generation, as larger households tend to make more trips but may share vehicles more efficiently. Income levels affect both trip generation and mode choice, with higher-income households typically owning more vehicles and making more trips.

UN estimates indicate that by 2030, there will be around 4.9 billion people living in urban areas worldwide, and by 2050, about 70% of people will be urban residents. This ongoing urbanization trend has profound implications for traffic growth, particularly in rapidly developing regions.

Economic Development and Employment

Economic activity drives traffic growth through multiple mechanisms. Employment levels directly affect commuting patterns, with job growth typically leading to increased traffic volumes during peak periods. Commercial and industrial development generates freight traffic as well as trips by employees and customers.

Economic prosperity generally correlates with higher vehicle ownership rates and increased discretionary travel. However, economic downturns can temporarily reduce traffic growth or even cause declines in traffic volumes. Research shows employment, population, and fuel price forecasts often lead to inaccurate traffic predictions. This highlights the challenge of forecasting economic variables that influence traffic.

Land Use Patterns and Urban Development

The spatial arrangement of land uses fundamentally shapes travel patterns and traffic volumes. Sprawling, low-density development typically generates more vehicle trips and longer trip lengths than compact, mixed-use development. The separation of residential areas from employment centers, shopping, and services increases travel distances and automobile dependence.

Conversely, mixed-use development that integrates residential, commercial, and employment uses can reduce trip lengths and enable more walking, cycling, and transit use. Transit-oriented development concentrated around high-quality transit stations can significantly reduce automobile traffic compared to conventional suburban development patterns.

Major developments such as shopping centers, office parks, or residential subdivisions can substantially affect traffic on nearby roadways. Site-specific traffic impact studies analyze these effects and often require developers to contribute to transportation improvements to accommodate the additional traffic their projects generate.

Transportation Policies and Infrastructure Investment

Transportation policies and infrastructure decisions significantly influence traffic growth patterns. Adding roadway capacity can accommodate growth but may also induce additional traffic. A literature review of several studies focused on induced demand found that between 50–100% of new roadway capacity is often absorbed by traffic within three or more years. This induced demand phenomenon means that capacity expansion alone rarely solves congestion problems.

Projection accuracy suffers from inadequate consideration of induced demand. Research shows new roadway capacity becomes 50-100% filled with induced traffic within just three years. Standard modeling procedures still largely exclude this phenomenon. This represents a significant challenge for traffic forecasting and infrastructure planning.

Pricing policies such as tolls, parking fees, and congestion charges can influence travel behavior and traffic volumes. Transit investments provide alternatives to automobile travel and can reduce traffic growth on parallel roadways. Complete streets policies that accommodate multiple modes can shift some trips from automobiles to walking, cycling, or transit.

City transportation policies often prioritize walking, bicycling, and transit. In some cases, cities aim to achieve explicit mode share targets to reduce dependence on single occupancy vehicle use. Meeting these aggressive goals and targets will require a shift in both infrastructure investment and traveler behavior.

Technological Advancements and Changing Travel Behavior

Technological changes are reshaping travel patterns in ways that challenge traditional forecasting approaches. Telecommuting and remote work reduce commute trips, a trend dramatically accelerated by the COVID-19 pandemic. Changes in travel patterns—like those from the pandemic and flexible work arrangements—have made long-range traffic forecasting trickier.

E-commerce and online shopping reduce trips to retail stores while increasing delivery vehicle traffic. Ride-hailing services like Uber and Lyft have added vehicle miles traveled in many urban areas. Autonomous vehicles, when they become widespread, could dramatically alter travel patterns by changing the cost and convenience of automobile travel.

Electric vehicles are changing the economics of vehicle operation, potentially encouraging more driving due to lower per-mile costs. Shared mobility services including car-sharing and bike-sharing provide alternatives to private vehicle ownership and may reduce traffic in some contexts.

Fuel Prices and Transportation Costs

Fuel prices significantly influence travel behavior and traffic volumes. Higher fuel prices typically reduce discretionary travel and encourage mode shifts to transit, carpooling, or more fuel-efficient vehicles. Lower fuel prices tend to increase vehicle miles traveled and can encourage purchases of larger, less efficient vehicles.

However, forecasting fuel prices over the 20-30 year horizons typical for transportation planning is extremely difficult. Price volatility and uncertainty about future energy markets complicate traffic forecasting efforts that attempt to account for fuel price effects.

Environmental and Climate Considerations

Growing awareness of climate change and air quality concerns is influencing transportation policies and potentially affecting traffic growth patterns. Many jurisdictions have adopted goals to reduce greenhouse gas emissions from transportation, which may require limiting traffic growth or shifting travel to lower-emission modes.

These growth estimates also directly shape air quality assessments, noise studies, and energy consumption calculations – which change drastically based on expected traffic levels. The environmental impacts of traffic make accurate forecasting important for environmental planning and regulatory compliance.

Practical Application of Traffic Forecasting Methods

Selecting and applying appropriate forecasting methods requires careful consideration of project characteristics, data availability, and resource constraints. Different project types and contexts call for different approaches.

Selecting the Appropriate Forecasting Method

Depending on scope and complexity of the analysis, different future methodologies are needed from simple historical trends to complex travel demand models. The selection process should consider several factors:

  • Project Scope: Large-scale regional projects typically warrant comprehensive travel demand modeling, while smaller projects may be adequately served by simpler trend-based methods
  • Planning Horizon: Short-term forecasts (5 years or less) can often rely on trend analysis, while longer horizons benefit from more sophisticated modeling
  • Data Availability: The availability of historical traffic counts, socioeconomic data, and existing travel demand models constrains method selection
  • Resource Constraints: Budget, staff expertise, and time limitations affect which methods are feasible
  • Required Accuracy: Projects with greater consequences for errors (such as major infrastructure investments) justify more rigorous forecasting methods
  • Local Context: Rapidly growing areas, areas with major planned developments, or areas with changing transportation systems may require more sophisticated approaches

The growth rate is a critical component needed to forecast traffic volumes. It can be determined using various methods. All methods listed in this section can be used to determine growth rate. The growth rate may vary from method to method, so it is recommended that multiple methods be checked for validation purposes.

Developing Growth Rates from Multiple Sources

Best practice in traffic forecasting involves developing growth rates from multiple sources and comparing the results. This triangulation approach helps identify outliers and increases confidence in the final forecast. Sources for growth rate development include:

  • Historical Traffic Trends: Analysis of traffic count data over 10-20 years or more
  • Travel Demand Models: Regional or statewide models that incorporate land use and socioeconomic forecasts
  • Demographic Projections: Population and employment forecasts from state demographers or planning agencies
  • Adjacent Projects: Traffic studies for nearby projects that may have analyzed similar corridors
  • Peer Comparisons: Growth rates from similar corridors or communities

When growth rates from different sources diverge significantly, analysts must investigate the reasons and exercise professional judgment in selecting appropriate values. Documentation of the rationale for growth rate selection is essential for transparency and future reference.

Accounting for Uncertainty in Forecasts

All forecasts are subject to uncertainty. It is risky to design a road facility to a precise future condition given the uncertainties in the forecasts. There are uncertainties in both the probable growth in demand and the available capacity that might be present in the future.

Slight changes in the timing or design of planned or proposed capacity improvements outside of the study area can significantly change the amount of traffic delivered to the study area during the analytical period. Changes in future vehicle mix and peaking can easily affect capacity by 10 percent. Similarly, changes in economic development and public agency approvals of new development can significantly change the amount of future demand.

To address forecast uncertainty, analysts should:

  • Develop Multiple Scenarios: Create low, medium, and high growth scenarios to bracket the range of possible futures
  • Conduct Sensitivity Analysis: Test how forecast results change with different assumptions about key variables
  • Build in Flexibility: Design infrastructure that can be adapted or expanded as actual conditions become clear
  • Monitor and Update: Regularly compare actual traffic to forecasts and update projections as needed
  • Document Assumptions: Clearly state all assumptions underlying forecasts so they can be revisited and revised

It is good practice to explicitly plan for a certain amount of uncertainty in the analysis. This level of uncertainty is the purpose of sensitivity testing.

Validation and Post-Construction Evaluation

A recent study investigated the postconstruction accuracy of traffic forecasts and revealed that traffic on roads in urban settings (arterials and collectors) were typically overestimated by a significant amount. This finding highlights the importance of validating forecasting methods and learning from past performance.

While trends indicate that traffic volumes have leveled off or even decreased over the past 10 years in jurisdictions throughout the United States, traditional forecasting substantially overestimates the potential for traffic growth. This systematic bias suggests that forecasting methods need to evolve to reflect changing travel patterns and policy priorities.

Traffic forecasting must grow beyond simple mathematical formulas as we look toward 2025 and beyond. This growth needs actual post-construction validation data. It should apply appropriate growth models based on network maturity and line up projections with broader policy objectives.

Agencies should establish programs to compare forecasts with actual post-construction traffic volumes. This feedback loop enables continuous improvement of forecasting methods and helps calibrate models to local conditions. Systematic evaluation of forecast accuracy can identify which methods work best in different contexts and reveal systematic biases that need correction.

Challenges and Limitations in Traffic Forecasting

Despite advances in data collection and modeling techniques, traffic forecasting remains challenging. Understanding these limitations helps set realistic expectations and guides appropriate use of forecasts.

The Fundamental Challenge of Predicting the Future

Traffic forecasting attempts to predict human behavior and societal trends decades into the future—an inherently uncertain endeavor. Unexpected events such as economic recessions, pandemics, technological disruptions, or policy changes can dramatically alter travel patterns in ways that forecasts cannot anticipate.

Projected and actual traffic volumes often don’t match up. Traditional forecasting methods struggle to keep pace with recent trends. Urban road traffic predictions miss their targets by a lot. This gap between forecasts and reality reflects the difficulty of predicting complex, evolving systems.

Data Limitations and Quality Issues

Forecasting quality depends fundamentally on data quality. Historical traffic counts may have gaps, errors, or inconsistencies. Socioeconomic data used in travel demand models may be outdated or unavailable at appropriate geographic scales. Land use forecasts that drive travel demand models are themselves uncertain predictions.

Traffic counts are usually collected through traffic counter/radar stations which only cover a small part of the road network. This limited coverage means that many roadways lack the historical data needed for trend-based forecasting, requiring estimation methods that introduce additional uncertainty.

Model Limitations and Assumptions

All forecasting models rely on simplifying assumptions about travel behavior and system relationships. Travel demand models typically assume that historical relationships between land use, socioeconomic characteristics, and travel behavior will continue into the future. This assumption may not hold if preferences, technologies, or policies change significantly.

Models also have technical limitations. They may not adequately represent all relevant factors, may have calibration errors, or may not capture non-linear relationships or threshold effects. The complexity of comprehensive models can make them difficult to validate and can obscure errors or questionable assumptions.

The Mismatch Between Forecasting and Policy Goals

Without doubt, the gap between forecasting practices and today’s transportation policy goals presents one of our biggest challenges. Many 3-year old cities have set ambitious targets to increase non-motorized transportation modes. Their traffic projections often use outdated assumptions of continuous vehicle volume growth. This mismatch weakens efforts to create greener, multimodal transportation systems.

Traditional forecasting methods that simply extrapolate historical trends may be inconsistent with policy goals to reduce automobile dependence, increase transit use, or promote active transportation. Forecasts that assume continued traffic growth can become self-fulfilling prophecies if they drive infrastructure decisions that accommodate and encourage more driving.

Long-Range Forecast Uncertainty

Transportation planners usually create growth projections that look 20 years past a project’s expected opening date. These long planning horizons, while necessary for major infrastructure projects, compound forecast uncertainty. Conditions 20-30 years in the future may differ dramatically from today in ways that are difficult to anticipate.

The further into the future a forecast extends, the less reliable it becomes. Yet infrastructure investments often have service lives of 50 years or more, requiring decisions based on highly uncertain long-range forecasts. This tension between the need for long-range planning and the limits of forecast accuracy is an inherent challenge in transportation planning.

Best Practices for Traffic Volume Growth Estimation

Despite the challenges, transportation professionals can improve forecast quality by following established best practices and learning from experience.

Use Multiple Methods and Compare Results

Relying on a single forecasting method or data source increases the risk of errors. Developing forecasts using multiple approaches—such as historical trends, travel demand models, and demographic projections—and comparing the results helps identify outliers and builds confidence in the final forecast. When different methods produce similar results, confidence increases. When results diverge, investigation of the reasons can reveal important insights.

Ground Forecasts in Quality Data

Invest in collecting and maintaining high-quality traffic count data. Establish permanent count stations to track long-term trends. Ensure that short-term counts are properly adjusted for seasonal and day-of-week variations. Validate and clean data to remove errors. The quality of forecasts cannot exceed the quality of the underlying data.

Document Methods and Assumptions

Thoroughly document all forecasting methods, data sources, assumptions, and decision points. This documentation serves multiple purposes: it enables review and validation of the forecast, provides a record for future reference, allows forecasts to be updated as conditions change, and facilitates learning from comparison of forecasts with actual outcomes.

It is recommended that the traffic growth rates from the various data sources be documented and submitted as part of the Traffic Projections Methodology Memo. This formal documentation ensures transparency and accountability in the forecasting process.

Consider Context and Local Conditions

Regional differences also play a key role in the forecasting process, as traffic patterns differ from dense urban centers to less developed rural areas. Generic growth rates or one-size-fits-all methods may not adequately reflect local conditions. Consider factors such as:

  • Local development patterns and plans
  • Regional economic trends and major employers
  • Transportation system characteristics and planned improvements
  • Local policies affecting transportation and land use
  • Geographic constraints and opportunities

Align Forecasts with Policy Objectives

Ensure that forecasting methods and assumptions align with adopted transportation and land use policies. If a community has goals to increase transit use, promote compact development, or reduce vehicle miles traveled, forecasts should reflect scenarios consistent with achieving those goals rather than simply extrapolating past trends.

This doesn’t mean forecasts should assume policy success without justification. Rather, forecasts should explore scenarios that reflect different levels of policy implementation and effectiveness, helping decision-makers understand what outcomes different policy choices might produce.

Regularly Update and Validate Forecasts

Forecasts should not be static documents. As new data becomes available, as conditions change, and as time passes, forecasts should be updated. Regular comparison of forecasts with actual traffic counts helps identify when forecasts are diverging from reality and need revision.

Establish a systematic process for monitoring actual traffic growth and comparing it with forecasts. When significant deviations occur, investigate the causes and adjust forecasting methods accordingly. This adaptive approach helps keep forecasts relevant and improves accuracy over time.

Communicate Uncertainty Clearly

Be transparent about forecast uncertainty and limitations. Present forecasts as ranges or scenarios rather than single-point predictions when appropriate. Clearly communicate the assumptions underlying forecasts and how changes in those assumptions would affect results. Help decision-makers understand that forecasts are tools for planning, not precise predictions of the future.

Traffic forecasting continues to evolve as new data sources, technologies, and methods become available. Several trends are shaping the future of traffic volume growth estimation.

Big Data and Real-Time Information

The proliferation of connected devices, GPS-enabled smartphones, and vehicle telematics is generating unprecedented volumes of traffic data. These “big data” sources provide much more comprehensive coverage of the road network than traditional traffic counters and can capture detailed information about travel patterns, speeds, and routes.

Analytics platforms are making it easier for transportation agencies to access and use this data for AADT estimation and traffic forecasting. Real-time data enables more frequent updates to traffic estimates and can help identify emerging trends more quickly than traditional annual count programs.

Advanced Modeling Techniques

Activity-based travel demand models represent an evolution beyond traditional four-step models. These models simulate the daily activity patterns of individual travelers, providing more behavioral realism and better ability to evaluate policies that affect travel choices. Agent-based models simulate the decisions and interactions of individual travelers, potentially capturing emergent behaviors that aggregate models miss.

Machine learning and artificial intelligence techniques are being applied to traffic forecasting with increasing sophistication. Deep learning models can identify complex patterns in large datasets and may improve forecast accuracy, particularly for short-term predictions. However, these methods require careful validation and their long-term forecasting performance is still being evaluated.

Integration of Multiple Data Sources

Future forecasting approaches will likely integrate diverse data sources including traditional traffic counts, GPS probe data, transit ridership, bike-share usage, pedestrian counts, and land use information. Fusing these multiple data streams can provide a more complete picture of travel patterns and how they’re changing.

Cloud-based platforms and improved data sharing are making it easier to access and integrate data from multiple sources. Standardized data formats and APIs facilitate data exchange between agencies and systems.

Scenario Planning and Adaptive Management

Given the uncertainty inherent in long-range forecasting, transportation planning is increasingly embracing scenario planning approaches. Rather than trying to predict a single future, scenario planning explores multiple plausible futures and develops strategies that perform well across different scenarios.

Adaptive management approaches recognize that the future is uncertain and build flexibility into plans and infrastructure. Rather than designing for a single forecast, adaptive approaches create options to adjust as the future unfolds and actual conditions become clear.

Focus on Multimodal Planning

Traditional traffic forecasting focused almost exclusively on automobile traffic. Future approaches will increasingly need to forecast demand for all modes—walking, cycling, transit, and emerging modes like e-scooters and ride-hailing—and understand how these modes interact.

Multimodal forecasting requires different data and methods than automobile-focused approaches. It also requires understanding how investments in one mode affect demand for others and how mode choice responds to changes in service quality, cost, and convenience.

Conclusion

Estimating traffic volume growth is both a science and an art, requiring technical expertise, professional judgment, and understanding of local context. While numerous methods and tools are available—from simple trend analysis to sophisticated travel demand models—no approach can perfectly predict the future. The key to effective traffic forecasting lies in understanding the strengths and limitations of different methods, using multiple approaches to validate results, grounding forecasts in quality data, and clearly communicating uncertainty.

Traffic forecasting is the life-blood of transportation infrastructure planning worldwide. Over the last several years, forecasting methods have changed by a lot, and they still determine how we spend billions on infrastructure each year. The stakes for getting traffic growth projections right have never been higher as we head into 2025.

As transportation systems become more complex, as new technologies emerge, and as policy priorities evolve, traffic forecasting methods must continue to adapt. The integration of big data, advanced modeling techniques, and scenario planning approaches offers promise for improving forecast accuracy and usefulness. However, the fundamental challenge remains: making informed decisions about long-lived infrastructure investments based on inherently uncertain predictions about future travel demand.

Success in traffic forecasting requires not just technical proficiency but also humility about the limits of prediction, transparency about assumptions and uncertainty, and commitment to learning from experience. By following best practices, leveraging new data sources and methods, and maintaining a critical perspective on forecasting limitations, transportation professionals can develop traffic volume growth estimates that effectively support planning and decision-making.

For those seeking to deepen their understanding of traffic forecasting methods, the Federal Highway Administration provides extensive guidance and resources. The Transportation Research Board publishes research on forecasting methods and their application. State departments of transportation often maintain traffic forecasting guidelines and data resources specific to their jurisdictions. Professional organizations such as the Institute of Transportation Engineers offer training and networking opportunities for transportation professionals working on traffic analysis and forecasting.

As we look to the future, the importance of accurate traffic volume growth estimation will only increase. Growing populations, evolving technologies, climate concerns, and changing travel preferences all create both challenges and opportunities for transportation planning. By continually improving forecasting methods and applying them thoughtfully, the transportation profession can better serve communities and create transportation systems that meet future needs efficiently, sustainably, and equitably.