Projected climate-driven changes in pollen emission season length and magnitude over the continental United States

Anemophilous pollen, while crucial for plant reproduction and gene dispersal, significantly impacts climate and human health, triggering allergic diseases like hay fever and asthma. Up to 30% of the global population suffers from pollen-induced respiratory allergies, imposing substantial economic burdens. Pollen emission is strongly influenced by meteorological factors, with temperature significantly affecting the timing of pollen seasons and precipitation impacting both short-term and long-term emissions. Warmer temperatures have historically led to earlier spring pollen seasons and delayed late-season pollen. Prolonged pollen seasons have been observed for various plant taxa. Increasing atmospheric CO2 concentrations can enhance pollen production through fertilization and increased allergenic protein content. While past studies have shown pollen phenology's responsiveness to climate change, large uncertainties remain due to sparse pollen observations in both space and time. Previous research often focuses on limited spatial or temporal scales or only includes individual taxa or a small subset of allergenic pollen. The diverse impact of climate drivers on pollen emission across various vegetation types necessitates a comprehensive, taxa-specific approach. This study addresses these limitations by employing a continental-scale pollen emission model, using Coupled Model Intercomparison Project version 6 (CMIP6) future climate data for two emission scenarios (SSP245 and SSP585), to project changes in pollen emissions for thirteen prevalent airborne pollen taxa in the continental United States. The study also assesses the influence of increased CO2 concentrations and land cover changes on pollen production.

Existing literature demonstrates a strong link between pollen emission and meteorological conditions, particularly temperature and precipitation. Temperature influences the timing of pollen seasons, with warmer temperatures resulting in earlier onset for spring-flowering taxa and delayed onset for late-flowering taxa. Studies have also documented prolonged pollen seasons. Precipitation's impact is complex, with heavy rainfall reducing pollen concentrations but long-term changes potentially altering pollen production. The influence of increasing atmospheric CO2 on pollen production is well-established, with laboratory studies indicating significant increases in pollen and allergen production for various species. However, uncertainties remain regarding the real-world impact of CO2 increases. Past observation-based studies on pollen phenology are typically limited by spatial and temporal constraints. Continental-scale studies have limitations, often encompassing only a subset of allergic pollen taxa or focusing on limited climate drivers. This study addresses these limitations by using a more comprehensive modelling approach.

The study utilizes the Pollen Emissions model for Climate Models (PECM) to simulate daily pollen emissions for thirteen major pollen taxa across the continental US. The model incorporates land cover, temperature, precipitation, and CO2 concentrations as drivers. The model calculates daily pollen emission flux as a function of land cover fraction, annual pollen productivity, phenological factors, precipitation effects, and a CO2 factor. Pollen phenology is modeled using a Gaussian distribution, with the start and end dates of the pollen season linearly related to previous-year average temperature. Annual pollen production is scaled to a temperature-dependent factor based on the observed relationship between annual total pollen counts and previous-year average temperature. A precipitation factor accounts for pollen scavenging during high rainfall events. The impact of CO2 is modeled by doubling the pollen production factor for the SSP585 scenario. The model employs both taxa-specific and plant functional type (PFT)-based land cover data. Climate data are obtained from fifteen CMIP6 models, regridded to match the model's spatial resolution. Pollen emissions are simulated for both historical (1995-2014) and end-of-century (2081-2100) periods under SSP245 and SSP585 scenarios. A sensitivity analysis using the Morris method is conducted to evaluate model uncertainties, focusing on parameters related to pollen production and phenology. The analysis identifies parameters most impacting the simulated pollen amount, showing temperature-dependent pollen production parameters as the primary drivers. For taxa with temperature-sensitive pollen season duration, phenology parameters also play a significant role.

Under the SSP585 scenario, the study projects significant changes in pollen emission phenology and magnitude. The pollen season is projected to start earlier (up to 40 days) and become longer (up to 19 days) due to temperature increases. Three categories of phenological shifts are identified based on the temperature response of individual taxa. Changes in maximum daily pollen emission are driven by the interplay of temperature effects on phenology and pollen production. Warmer temperatures increase annual pollen production for some taxa while decreasing it for others. These effects, along with the impacts of precipitation, result in maximum daily pollen emission changes ranging from -35% to 40%. Temperature and precipitation alter maximum daily pollen emissions (-35% to 40%), while considering both effects with CO2 increases the maximum emission by up to 200%. Regional variations in pollen emission changes are observed, depending on the dominant vegetation taxa and their responses to climate change. For instance, in the Northeast, some deciduous broadleaf taxa show reduced maximum daily pollen emissions, while others experience increases. The convergence or divergence of individual taxa pollination impacts the overall maximum daily pollen emissions. The annual total pollen emission is projected to increase by 16-40% across the United States under the SSP585 scenario, largely driven by changes in pollen production and precipitation. When considering the effects of increased CO2 on pollen production, the annual total pollen emission increases up to 250%. Land cover change, based on PFT-based model, shows smaller impacts on pollen emissions compared to climate and CO2 changes. The projected changes in pollen seasons and emissions are consistent with past observations.

The findings underscore the significant impact of climate change, particularly temperature and CO2, on pollen emissions across the continental US. The model projections highlight the importance of considering both phenological shifts and changes in pollen production to accurately assess future pollen emission levels. Regional differences in vegetation composition and the responses of individual taxa to climate change drive regional variations in projected pollen emission changes. The inclusion of CO2 effects significantly magnifies the projected increase in pollen emissions, indicating a potential for substantial increases in allergenic pollen exposure in the future. The relatively smaller impact of land cover change suggests that climate drivers may be more dominant in the near term. The consistency between the model projections and historical observations validates the model's predictive capabilities.

This study projects substantial increases in pollen season length and magnitude across the continental US under future climate scenarios, particularly with elevated CO2. The findings highlight the complex interplay of temperature, precipitation, CO2, and land cover changes in shaping future pollen emissions. The large projected increases in pollen pose significant implications for public health, potentially exacerbating allergic diseases like hay fever and asthma. Future research should focus on improving model parameterizations, particularly regarding pollen production and CO2 effects, and incorporating more precise land cover change data to refine projections. The research demonstrates the urgent need for proactive strategies to mitigate the health impacts of increased pollen exposure.

The study acknowledges several limitations. The model's parameterizations rely on geographically limited historical pollen data, potentially affecting the accuracy of simulations for natural forest emissions. Uncertainty exists in the parameterization of CO2 effects on pollen production, owing to limited experimental data. The lack of gridded taxa-specific land cover change data limits the assessment of land cover change effects on pollen emission, and the PFT-based model provides only a broad estimate. Future research needs to address these uncertainties by expanding pollen observation efforts and conducting more extensive studies on the interplay between CO2, climate, and pollen production. Improved and more spatially explicit data on future land cover will help to improve the reliability of pollen emission predictions.