The Interactive Qualifying Project (IQP) is a project which requires students to work in interdisciplinary teams to solve a problem related to the intersection of science and society. Teams are usually comprised of students from different majors with different skills. The project is designed to expose students to global issues, help them learn to make critical decisions, work within a team and appreciate the societal context of their work.

We completed our project in two seven-week long periods. During the first period, the preparation term, we developed our initial proposal for our project. During the second period, we are currently executing our proposal and beginning to draft our final report.

To learn more about the IQP please visit:

https://www.wpi.edu/academics/undergraduate/interactive-qualifying-project

As pollen allergies are rising in Romania, allergists are noticing the lack of local research in this field. Our project aimed to create a software tool to optimize data analysis and discover correlations between pollen allergies and environmental factors to assist the research team at the Allergology Laboratory at Colentina Clinical Hospital and Carol Davila University in Bucharest, Romania and raise public awareness of pollen allergies. We coded a software tool using Python, gathered data previously collected in Bucharest on pollen counts, chemical pollution and climate, developed a survey to collect lifestyle data, and calculated Spearman’s correlation between factors. We determined that grass pollen levels were only correlated with O₃ (ozone) levels, while Ambrosia pollen correlated with humidity, precipitation and CO (carbon monoxide) levels. Using data from Google Trends we discovered that the frequency Romanians searched the words “Ragweed” and “Allergy” correlated strongly with various environmental factors. We compared our results to studies from other countries and found the majority were consistent with our findings. Most of our lifestyle survey respondents were from urban areas, experienced allergy symptoms during the Ambrosia season, and experienced allergy symptoms while at work. We concluded pollen allergies are more likely to worsen during dry, warm periods. We saw ozone to be the most inflammatory pollutant. Additional data should be collected over time to confirm the correlations and better understand Romanian lifestyle choices. More development of the pollen collection program is needed to predict and publicize daily pollen forecasts in the future.

Pollen allergies have been a problem around the world, affecting millions of people globally (Montagnani et al., 2017) and having a negative impact on the economy (Walker et al., 2016). To address the pollen allergy problem, many countries in Europe and North America have done intensive research and made efforts to help diminish its effect. In Romania, pollen allergies have increased both in quantity and severity. However, until 2014 there was no data collected to study this issue in Bucharest, the capital of Romania, which has been noticeably impacted. Allergists in Bucharest have seen a significant yearly increase in the number of patients with pollen allergies over the last few years and the symptoms they experience can significantly damage their respiratory health (Leru et.al 2018). It is important to make the public aware of the changing pollen levels and provide methods to predict when their allergy symptoms may worsen so that they can prepare and limit the effects on their health.

Allergenic pollen comes from many species. The most common groups are trees, grasses, and weeds, collectively producing pollen almost every month throughout the year. Among all species, Ambrosia (commonly known as “ragweed”) leads in pollen count and allergenicity (Leru et al., 2019). Our project focused on investigating the effects of climate change, air pollution, and human activities on the pollen counts of grasses and ragweed. Climate change increases the average global temperature and alters season durations, lengthening the pollen production period (D’Amato et al., 2010). Urbanization also plays an important role, as the number of pollen allergy patients reported from urban areas triples those from rural areas (Cvetkovski et al., 2018). The city provides beneficial growing conditions for invasive species like ragweed (Deutschewitz et al., 2003) and high levels of air pollution worsen allergy symptoms. Lastly, lifestyle choices can have an impact on pollen levels and personal development of an allergy. For example, by using more cars, we are adding to the ambient pollutant levels. Other factors such as medical history, household conditions, diet, whether a person grew up in an urban or rural area also impact the pollen allergy development.

The goal of our project was to help assess the situation of pollen allergies, climate and air pollution in Romania and how they are related, as well as raise awareness of this situation to advocate for better policies and further research. We collaborated with Dr. Polliana Leru and her research team at Carol Davila University and the Allergology Laboratory at Colentina Clinical Hospital to help optimize their data organization and analysis processes to determine which meteorological, environmental and human activity factors affect the rise in pollen allergies.

During our project, the global outbreak of COVID-19 occurred. Travel restrictions and health advisories prevented our group from traveling to Bucharest to carry out our project. However, due to the importance of global climate change and the related medical issues it is causing, including seasonal allergies, we committed ourselves to carry out the project remotely, through electronic communications. Our collaborator, Dr. Leru was equally committed and eager to continue the project, on top of her responsibility as a doctor at a hospital that treated COVID-19 patients.

To accomplish our goal we first obtained and organized the data, then developed our data analysis tool, performed data analysis, gathered missing lifestyle data and finally devised a strategy to raise awareness about the rise in pollen allergies in Romania. In addition, we developed a project website, where we have posted our most important findings, linked here: https://sites.google.com/view/wpi-iqp/home.

The data used in our project included many factors such as pollen counts, meteorological factors, air pollution levels and public interest. The pollen data was gathered from the Allergology Laboratory at Colentina Clinical Hospital, where it has been collected using a Burkard pollen trap since 2014. The meteorological data came from the National Meteorological Agency in Romania and included variables such as temperature, relative humidity, wind speed, and wind direction. The chemical air pollution data came from the National Air Quality Monitoring Network stations closests to Colentina Clinical Hospital. Lastly, we used Google Trends to gather data about Romania’s public interest in specific search terms over time using the terms “Allergy” and “Ragweed” as an indicator of the number and severity of allergy cases in Romania at different times over the years.

Our data analysis tool is called the “Correlation Machine”. It was coded using Python and was designed to perform three main tasks: manage data, calculate Spearman’s correlation coefficients and generate graphs with regression models. Pictures of each tab can be found in section 4.1 of this report. Our source code files can be accessed and downloaded from Github: https://github.com/trschaeffer/PollenTool The link to our executable file can be found here: https://drive.google.com/open?id=1HCP7sD6GoTpnJ_JSF1R0VvUTVGMOWXey

The first section, or tab, of the Correlation Machine is where data management is performed. The user has the ability to load files into the tool. All files that are loaded are combined into one large file known as the “master”. In this tab, the user can also manually add a single data point to the master file and calculate the total daily pollen levels.

The second tab is where all of the calculations are performed between any two categories that the user selects. There are multiple filters that can be optionally applied to the correlation, such as choosing pollen season dates, calculating monthly or yearly averages, or only including data within a certain time period. The user can apply any combination of these filters to the data. After clicking “Calculate”, Spearman’s correlation coefficient, the p-value (or probability of the correlation’s true existence) and a list of the applied filters will appear in a row in the table at the bottom of the tab. This table can be saved as an Excel spreadsheet for future reference.

The third tab is used for generating graphs. Up to six variables can be graphed at a time. The user can select to plot each variable on the left or right y-axis for better visibility of the lines. This tab also allows the user to apply the same data filters as the calculations tab. In addition to those filters, regressions can be plotted by choosing a regression type and entering a year in the future that the line will be extended too. After clicking “Generate Graph”, a graph will appear in a new window, which can be analyzed and saved to the computer.

The final tab is called the “Help” tab. In this tab there are links that go directly to Youtube videos with demonstrations of how to perform specific tasks. There is also a link to our user’s guide which is a written version of all of the videos put together.

The Correlation Machine was designed to store all available data, perform analysis, generate graphs, and can be adaptive to new data in the future. With an intuitive User Interface and video instructions, users with any background can learn the tool quickly. Through the development of the tool, we hope to help optimize the current and future data analysis.

Using the Correlation Machine, we generated results to better understand the interactions between pollen levels, pollen symptoms, and various external factors. We found that the total pollen levels in the air are higher on hot, dry days. We did not find any correlation between total pollen levels and PM₁₀ that we had expected. The only types of pollen that correlated with PM₁₀ from May until October were Artemisia and Ambrosia, which we have learned from Dr. Leru is not often confused for PM₁₀.

The strongest correlations we found between ragweed pollen levels and meteorological factors were with relative humidity and precipitation during the full pollen season. There is a negative correlation with both of these factors; when there is less moisture in the air, the pollen count is higher. Ragweed pollen levels were also positively correlated to carbon monoxide levels (CO), although there has not been much biological research explaining the relationship between these factors.

The most significant factors correlated to grass pollen levels were ozone (O₃) and solar radiation, specifically during the peak grass season from April 25 until the end of June. Both showed a positive correlation. While the strongest correlations with O₃ were found looking at yearly averages, the monthly averages were the only scenario with a strong correlation and a statistically significant p-value. Even though we cannot show causation between these factors and grass pollen levels, this correlation suggests that there may be some link between the O₃ levels, solar radiation and grass pollen over time.

We were able to find many factors that correlated with the public’s interest in searching ‘Allergy’ and ‘Ragweed’. Again the theme of hot, dry, sunny days being the worst for pollen allergies was reinforced through correlations with temperature, humidity, and solar radiation. Ozone was one of the strongest correlated pollutants, leading us to believe that it is the strongest lung irritant. Other pollutants did not seem to have much of an effect, though they may be an underlying cause of pollen allergies, and many of them are involved in the ozone production process.

After compiling our results together from the correlation machine, we chose to compare these results to studies conducted in other countries to see if our findings were similar or different.

First we examined a study conducted in Northern Italy conducted from 1981 to 2007. This study compared grasses and other pollen with temperature, humidity, precipitation, wind speed and solar radiation. They found no significant change in grass pollen over time, while temperature and humidity linearly increased. This resulted in a weak correlation between grasses and temperature, while grasses correlated with humidity (Ariano et al., 2010). In our results we also saw the same constant trend with grass pollen and increasing trends with temperature and humidity. Because this study has significantly more data than our project, and we discovered similar trends and correlations, we expect this trend to hold true over time as more data becomes available.

We also compared our results to studies conducted in the United States. In one study, conducted from 1995-2009, results showed that the length of the Ambrosia pollen season correlated directly to the number of frost free days and latitude. These findings suggest that areas at higher latitudes are more affected by climate change, shown by the increasing length of the frost free periods each year. For latitudes above 44^oN, the season increases as much as 13-27 days in 2009 compared to 1995. Even though analysis with temperature within our project was inconclusive, Romania has an average latitude of 45.9432° N, therefore, it is likely that over the years, with more data collected, one will also observe the increase in Ambrosia season duration due to climate change (Ziska et al., 2010).

Based on the correlations we made and findings from other research, we did a prospective analysis to produce future predictions for two major events taking place in Bucharest, climate change and pollution awareness. As the climate continues to warm, more pollen will be generated. Hot, dry days were shown to be the worst for pollen allergy symptoms. We observed lower average humidity and higher average temperatures in our analysis, meaning Bucharest will experience more of these days and worsening allergy symptoms. Regarding pollution, Bucharest’s PM_2.5 and PM₁₀ level, both tend to peak well above the European Union legal limits. Without any intervention, these chemicals will continue to cause lung irritation, which can lead to the development of pollen allergies in more people. Fortunately, pollution activism has been brought to attention by independent companies such as Airly and Aerlive, and hopefully Romania will begin to take action on reducing its pollution level, possibly preventing many people from developing allergies.

Lifestyle is another factor associated with pollen levels and symptoms. We distributed our lifestyle survey to an Ambrosia sufferer Facebook group in Romania. We received 92 responses and 87% of them are affected by pollen allergies (including ragweed, trees, and grasses). 47% are from Muntenia region where Bucharest is located and 63% reported to grow up in urban areas. This finding justifies the common assumption that urban areas are most affected by pollen allergies. The sufferers reported to experiencing their symptoms throughout the year, but especially during ragweed season, suggesting that ragweed sensitization is the most common of allergies. However, since people also have their symptoms in other months, we can conclude that many are affected by more than just ragweed. We also found that pollen allergies are not so much genetically driven, and instead are more likely to be impacted by environmental factors. This is evident as the majority of respondents said they did not have any family or personal history of respiratory diseases. Although it is true that the majority of Romanians choose cars as their primary mode of transportation, more studies are needed to determine the relationship between Romanian’s car usage and their allergies. Lastly, we found that allergy symptoms at work are surprisingly common. Majority (75%) of the respondents reported to experience their symptoms at work, which can affect their productivity, the quality of their work, and sometimes even their own safety. This finding especially stressed the importance of doing more research to help people cope with the effects of allergies.

To help raise public awareness of pollen allergies in Romania, we designed an educational pamphlet to be distributed in the hospital or online. The pamphlet contains basic information about allergies, factors that contribute to the spread of pollen, common symptoms, and possible treatments or ideas for prevention. In addition, we created a pollen allergies Facebook page where we added short informational posts over the course of our project to provide more specific information about allergies. The pamphlet has also been posted to this page to make it available to a wider audience. We also linked our survey to this page to gather more responses (https://www.facebook.com/polleninromania/).

After the completion of our project, there are a few things we believe could be added or expanded on in the future. Our biggest recommendation is for the continuation of data collection, as this will strengthen the confidence in our results so far, allow more future trends in Bucharest to be determined, and help find more correlations in places where our current data set is limited. Additionally, we believe that calculating and publishing pollen forecasts in Bucharest, adding health data as a factor, collecting more survey results to a statistically significant level, and investigating whether there is a link between COVID-19 and pollen allergies would be useful additions to our project in the future.

At the completion of our project, we compiled and presented all relevant information about pollen allergies from global and local sources, developed a “Correlation Machine” for ease of data analysis, and proved that Romania is observing a similar trend to other countries that are heavily impacted by pollen. Our project team, along with our collaborator, Dr. Leru and her research team believe that with climate change, urbanization, and the immediate threat of COVID-19, it is more important now than ever before to gather more research in Romania regarding pollen allergies in order to protect the health of the public.