Forecasting Climate-Driven Bitumen Performance Requirements in Iran Using Prophet Time Series Modeling

Authors:

Mohammad Mehdi Dadaei, Pouria Hajikarimi, & Fereidoon Moghadas Nejad

Introduction

Bitumen is a critical material in road construction, and its performance is highly sensitive to temperature changes. As climate change intensifies, understanding the effects of rising temperatures on bitumen behavior becomes crucial—especially in countries like Iran that experience extreme temperatures. The Performance Grading (PG) system, widely used to classify bitumen based on rheological properties under specific climate conditions, is directly influenced by ambient temperature trends.

This study uses Prophet, a powerful time series forecasting model developed by Facebook, to predict air temperature trends for 13 key Iranian cities from 2025 to 2060. The goal is to evaluate how anticipated climate changes may alter the performance grade requirements for bitumen across different regions. Additionally, the study explores environmental and economic aspects, integrating climate forecasts with assessments of additive use and local material availability to propose sustainable and cost-effective strategies for future asphalt design.

2.1 Study Area and Climate Classification

Iran’s vast and diverse landscape leads to varied climatic conditions that directly impact asphalt pavement design. The study focuses on four representative cities across distinct climate zones:

•Tehran: Semi-arid, with cold winters and hot summers; highly urbanized and trafficked.

•Bandar Abbas: Hot and humid, located on the southern Persian Gulf coast, with consistently high year-round temperatures.

•Tabriz: Cold semi-arid, in the northwest; ideal for assessing low-temperature pavement behavior.

•Ahvaz: Hot desert, in the southwest; extreme summer heat, among the country’s highest recorded.

These cities were chosen for their geographic spread, traffic density, economic significance, and climate diversity, enabling a comprehensive evaluation of temperature effects on bitumen performance.

The Köppen-Geiger climate classification was used to group Iran into three main climate categories:

•Hot and Humid (Group A): Southern coastal areas like Bandar Abbas.

•Cold-Temperate (Group B): Northwestern areas like Tabriz.

•Arid and Semi-Arid (Group C): Central plateau and urban centers like Tehran and Ahvaz.

2.2 Data Collection and Preprocessing

Historical daily temperature data (1990–2020) were sourced from the Islamic Republic of Iran Meteorological Organization (IRIMO), covering minimum, maximum, and average daily temperatures across key cities.

To prepare the data:

•Missing values were filled using linear interpolation or seasonal decomposition (STL).

•Outliers were identified using z-score and IQR methods and corrected or removed.

•Daily mean temperatures were calculated as the average of minimum and maximum values.

The cleaned dataset was transformed into Prophet-compatible format, retaining seasonal anomalies (like El Niño) relevant to long-term projections. A 5-year rolling average was applied to highlight climate trends, aligning with the 20–40-year service life window typical in pavement design.

The data were divided into the three climate macro-zones to enhance policy relevance. Time-domain decomposition was performed to extract seasonal and trend components, confirming the presence of annual seasonality and a consistent upward temperature trend.

Temperatures were converted from Celsius to Kelvin. Additionally, the averages of the seven hottest and coldest consecutive days per year were calculated to match with PG temperature thresholds, helping evaluate how future climate may shift bitumen grade requirements.

2.3 Forecasting Future Temperatures with Prophet

2.3.1 Model Rationale and Advantages

The Prophet model was selected over traditional methods (ARIMA) and machine learning models (LSTM) due to its:

•Transparency and interpretability (additive decomposition: trend + seasonality + error).

•Automatic detection of yearly/weekly cycles, critical for climate studies.

•Scalability for large datasets and long-term forecasts.

•Flexibility to handle nonlinear trends and abrupt shifts (changepoints, custom growth curves).

2.3.2 Model Setup and Parameters

For each city, Prophet was applied using 2000–2023 historical data. Key parameters:

•Trend: Linear growth with automatic changepoint detection.

•Seasonality: Yearly cycles using Fourier series (order = 10).

•Forecast horizon: January 1, 2024 – December 31, 2060.

•Outliers: Retained to reflect realistic extreme events.

Hyperparameters were tuned using Python’s prophet library with cross-validation on the last five years of data.

2.3.3 Validation and Performance

Model validation (2018–2023) used:

•Mean Absolute Error (MAE)

•Root Mean Squared Error (RMSE)

•R² Score

Results showed MAE between 0.9°C–1.6°C and R² > 0.85 across cities, confirming reliability for medium- and long-term projections.

2.3.4 Forecasted Trends (2024–2060)

Key projected increases in average annual temperatures:

•Tehran: +2.5°C by 2060, with intensified summer extremes.

•Bandar Abbas: +1.8°C, reinforcing its hot-humid profile.

•Ahvaz: +3°C+, raising concerns over bitumen softening and rutting.

•Tabriz: +1.6°C, crucial for low-temperature performance.

These projections are essential for updating bitumen binder specifications to maintain pavement durability under changing climates

2.4 Bitumen Performance Grading Framework

2.4.1 Overview of the Superpave PG System

The Performance Grading system, part of the Superpave methodology, defines bitumen grades based on:

•High-Temperature Grade (PG-XX): Resistance to rutting at peak pavement temperatures.

•Low-Temperature Grade (PG-YY): Resistance to thermal cracking at coldest pavement temperatures.

Example: A PG 64-22 binder is suitable between 64°C (high) and -22°C (low) pavement temperatures.

2.4.2 Linking Forecasts to PG Grades

Steps to connect Prophet forecasts to PG requirements:

1.Air-to-pavement temperature conversion: Using the LTPP Bind method, applying corrections (typically +2 to +8°C for high, -1 to -5°C for low).

2.PG grade estimation: Calculating the 98th percentile of high and 2nd percentile of low temperatures per year, mapped to PG thresholds.

3.Time horizon categorization:

•Present Day (2020–2023)

•Mid-Century (2040–2049)

•End-Century (2050–2060)

This approach assesses how climate-driven changes will shift bitumen grade requirements over time.

2.4.3 Observed and Projected PG Requirements (By City)

The following PG (Performance Grade) shifts were identified:

•Tehran:

            •Present: PG 64-10

            •2040s: PG 70-10

            •2050s–2060: PG 76-10

•Bandar Abbas:

            •Present: PG 70-16

            •2040s: PG 76-16

            •2050s–2060: PG 82-16

•Ahvaz:

            •Present: PG 76-10

            •2040s: PG 82-10

            •2050s–2060: PG 82-10 (with occasional PG 88 events in extreme years)

•Tabriz:

            •Present: PG 58-22

            •2040s: PG 64-22

            •2050s–2060: PG 64-16

These trends show a consistent shift toward higher-temperature PG grades, particularly in southern and central Iran, raising concerns about bitumen softening, rutting, and premature pavement failure unless advanced modifiers are adopted.

2.4.4 Implications for Bitumen Selection and Pavement Design

The shift in Pg grades implies:

•A growing need for polymer-modified and additive-enhanced bitumen to withstand elevated high temperatures (e.g., PG 76+ grades).

•The potential for performance mismatches if old specifications are used, especially in provinces that currently do not incorporate climate forecasts.

•Long-term pavement life may be compromised unless agencies adopt a dynamic PG assignment approach that evolves with projected climate data.

This sets the foundation for exploring additives like SBS, PPA, and CR in the next section, focusing on how chemical modification can extend the thermal operating window of base binders and maintain compliance with future PG demands.

2.5 Economic and Environmental Considerations of Additives in Bitumen Modification

The economic viability and environmental impacts of using additives like SBS (Styrene-Butadiene-Styrene), PPA (Polyphosphoric Acid), and CR (Crumb Rubber) are crucial in pavement engineering decisions.

•SBS (polymer modifier):

•Significantly improves mechanical properties and durability, extending pavement lifespan and reducing maintenance costs.

•However, it involves higher initial expenses, making it challenging for budget-constrained projects.

•PPA (chemical modifier):

•Offers an economically favorable option by enhancing high-temperature performance without extensive polymer addition.

•Improves rutting resistance and binder stiffness, while showing a lower carbon footprint compared to SBS.

•CR (Crumb Rubber, from recycled tires):

•Provides both waste management and bitumen enhancement benefits.

•Improves elasticity and temperature susceptibility, addressing environmental concerns around tire disposal.

•Challenges include variability in CR properties and complex processing that may affect performance consistency.

In summary:

•SBS = superior performance but higher cost.

•PPA = cost-effective enhancer with environmental advantages.

•CR = sustainability-friendly but needs careful application control.

2.6 Economic and Environmental Evaluation of Additive-Based Adaptation Strategies

As Iran faces increasingly severe climate conditions, asphalt binder strategies must balance performance, cost-efficiency, and sustainability.

Life-Cycle Cost Analysis (LCCA)

A 35-year LCCA using data on material prices, maintenance, and rehabilitation cycles shows:

•SBS and SBS+PPA binders, despite higher upfront costs, extend service life significantly, reducing total life-cycle costs (doubling maintenance intervals from ~5–6 to 10–12 years).

•CR binders are a cost-effective intermediate, providing moderate performance gains at lower added cost—especially suited for moderate climate shifts (e.g., Tabriz, Kermanshah).

•Unmodified binders become economically unsustainable in hot southern cities like Ahvaz and Bandar Abbas due to accelerated degradation.

Life Cycle Assessment (LCA)

Environmental assessments showed:

•SBS-modified binders have higher environmental load due to petrochemical and energy demands.

•PPA addition slightly reduces this impact by improving efficiency.

•CR binders show the best environmental profile, offering low GWP, low energy demand, and positive contributions via tire recycling—especially valuable in polluted cities like Tehran and Mashhad.

Policy and Market Considerations

To enable nationwide adoption:

•Incentivize CR-modified binders through subsidies and recycling support.

•Reduce import dependency on SBS by boosting domestic production or creating national reserves.

•Update performance grading specifications to reflect climate forecasts, using tiered standards matched to projected conditions.

3. Policy Recommendations for Climate-Responsive Asphalt Binder Strategies in Iran

Iran’s shift toward a climate-resilient road network needs integrated policies combining climate science, materials engineering, and economic planning.

3.1 Update National Performance Grading Standards

•Revise PG systems to integrate forecast-based temperature ranges up to 2060.

•Establish region-specific PG thresholds (e.g., PG 76-XX+ for Ahvaz/Bandar Abbas; PG 64-XX for Rasht/Urmia).

•Adopt outputs from climate models (Prophet, CMIP6, CORDEX) as official PG map inputs.

•Update these maps every 5–10 years.

3.2 Establish Additive Suitability Zones and Guidelines

•Define SBS-dominant zones (hot, high-traffic provinces: Khuzestan, Hormozgan, Kerman).

•Define CR-dominant zones (moderate climates + air pollution: Tehran, Tabriz, Kermanshah).

•Recommend PPA as a cost-reducing enhancer where budgets are constrained.

3.3 Incentivize Domestic Production and Recycling

•Invest in local SBS manufacturing to reduce costs and secure supply.

•Subsidize CR production via tire recycling programs.

•Establish national additive quality standards to ensure consistency.

3.4 Integrate Economic and Environmental Metrics

•Require life-cycle cost-benefit analysis (LCCA) for major projects.

•Mandate environmental impact reporting using standardized LCA metrics.

3.5 Build Capacity for Climate-Aware Engineering

•Launch training programs on climate-resilient materials and sustainability.

•Fund pilot projects and demonstration sites for additive testing.

•Create a centralized data platform to monitor pavement performance, additive use, and environmental impacts.

4. Conclusion

This study highlights the urgent need for Iran to reform asphalt binder grading and selection strategies, given projected climate shifts through 2060.

Key takeaways:

•Southern/central provinces will require high-temperature PG grades due to warming.

•No single additive solution fits all regions—SBS suits extreme conditions, PPA improves efficiency, and CR balances sustainability and performance.

•Region-specific strategies, climate-integrated specifications, and life-cycle evaluations are essential for long-term resilience.

By combining climate forecasts, materials science, economic analysis, and policy action, Iran can build a durable, cost-effective, and environmentally responsible transportation system for the coming decades.

This research serves as a practical guide for planners, regulators, and industry stakeholders to make adaptive, forward-looking decisions in Iran’s asphalt sector.