Five Questions: FAMU-FSU professor uncovers the impacts of hurricane flooding on mold growth

By: Kayla Cardenas | Published: September 11, 2024 | 12:09 pm | SHARE:

Atlantic hurricane season is nearing its peak, raising alarms for mold outbreaks triggered by flooding and the respiratory health issues to follow.

Ebrahim Ahmadisharaf, an assistant professor and researcher at the FAMU-FSU College of Engineering’s Resilient Infrastructure and Disaster Response Center, or RIDER, is shedding new light on the indirect effects of flood damage on residential buildings and human health.

His interdisciplinary team of hydrologic engineers and public health scientists explores how flooding after hurricanes contributes to mold growth in homes and the subsequent impact on indoor air quality and respiratory health.

Their research focuses on water-damaged homes in flood-stricken areas such as New Orleans, Baton Rouge, South Florida, New York City and Philadelphia, where they analyzed mold proliferation and its connection to respiratory illnesses. The team aims to uncover the key factors driving mold growth, such as flood depth and roof age, and their correlation with asthma and allergy symptoms. This study addresses critical knowledge gaps and offers insights into the long-term health effects of living in flood-impacted environments.

What is the relationship between flooding and mold growth?
Numerous studies have focused on predicting floods and assessing their impact on fatalities and structural damage. However, when it comes to the indirect and less visible consequences, like mold growth, there has been little quantitative research — particularly in measuring total mold spore counts.

Flood-related conditions, such as excessive dampness and humidity, create an ideal environment for mold to thrive, which can lead to significant respiratory health issues. Mold-related problems don’t necessarily require standing water in the building. Depending on the building materials, water can remain trapped, fostering conditions for mold growth. In severe cases, invisible mold can develop, which often goes unnoticed by residents.

We aimed to collect data on the factors that contribute to mold proliferation and assess its impact on respiratory health, particularly asthma and allergies. Our goal was to predict these outcomes using machine learning algorithms.

What is invisible mold, and how does it pose a threat without people realizing it?
When most people think of mold, they picture black spots on walls or around exhaust fans. But there are other types of mold that are invisible to the naked eye. These unseen mold spores can cause the same respiratory issues as visible mold, and in some cases, may lead to more prolonged health problems.

The challenge with invisible mold is that people often don’t realize it’s present, so they don’t take steps to remediate it. As part of our study on the mid-term impacts of mold growth, we interviewed residents to understand what actions they had — or hadn’t — taken to address mold and whether they had experienced any respiratory issues. This gave us insight into the long-term health effects associated with invisible mold.

How did you incorporate machine learning into your research, particularly with modeling work?
We employed a range of methods to gather data. First, we conducted online questionnaires and surveys to gather information on building characteristics, such as roof conditions and ventilation systems. Then, we visited the properties and collected indoor and outdoor air and dust samples.

After analyzing the samples in the lab, we identified more than 40 species of mold. In addition, we used hindcast data related to flood conditions, along with machine learning algorithms and public domain datasets. Before combining all the data, we took steps to validate it. For instance, we cross-referenced our flood models with stream gauge data and high-water marks.

We also gathered details from participants about the flood levels around their homes and how long the water remained. We then fed this data into machine learning models to predict total mold spore counts based on variables such as roof age, flood depth and ventilation.

The algorithms revealed key factors influencing mold growth, which we used to predict mold levels under certain conditions. In addition to our mold growth model, we developed a classification model to predict asthma exacerbations. By analyzing participant data on symptoms and hospital visits post-flood, the model helped identify the conditions under which asthma symptoms worsened. These insights can inform remediation and prevention efforts.

What is the concept of hindcasting, and how does it fit into this study?
Hindcasting is similar to forecasting but works in reverse — it recreates past events using models and historical data to predict future scenarios. It’s especially useful when we don’t have flood data from every affected building. Hindcasting helps fill those gaps and enhances the spatial coverage, providing insights into how flooding evolved, the maximum flood depth, how long the water remained and how it receded. This technique supports both historical analyses and future flood predictions, offering more complete temporal coverage across different locations.

What does the future of this research look like?
Several avenues could expand our current research. In terms of flood modeling, high-resolution models at the building level could enhance the accuracy of flood depth predictions. For flood characteristics, gathering more detailed data on flood duration, velocity and rate of rise would help validate models and improve the reliability of our conclusions. To better understand flood impacts, we could track data over time for each affected building, capturing both immediate and long-term effects.

While our current focus is on extreme hurricane-related flooding, smaller flood events, such as those caused by heavy rainfall, should also be studied. In some cases, minor flooding could have an equal or even greater impact due to prolonged wetness and saturated conditions from back-to-back storms.

More research in this area would help educate communities about flood preparedness for both major and minor events, provide useful insights for building design, and inform occupational health science and public health policy. We hope that our research efforts will generate more awareness about indoor mold proliferation and the impacts on respiratory health, especially in communities at high risk for allergy and asthma symptoms.