Seasonal Forecasting of Floods Using Climate Information
A new framework for flood prediction and risk management
About Me
Hello world!
I am Equisha Glenn, a NOAA EPP Center for Earth System Sciences and Remote Sensing Technologies (CESSRST) graduate research scholar attending the City College of New York.
As a CESSRST Civil Engineering PhD candidate, my research demonstrates the intersection of water resources, climate, and resiliency. My work focuses on understanding the influence of climate changes on water resources to adapt water management strategies for floods and drought. I aim to bridge the gap between, water resources, climate, and policy to help build a more resilient future for cities and society.
CESSRST Faculty Advisors:
Naresh Devineni, PhD | The City College of New York
Jorge E. Gonzalez, PhD | The City College of New York
NOAA Mentor:
Tom Smith, PhD | NOAA/STAR/SCSB & CICS/ESSIC, U. of Maryland
Fun fact...
I have been a NOAA EPP Scholar throughout my undergraduate, masters and PhD degree programs!
To learn more about opportunities for students including high school to PhD level programs visit the site:
Introduction & Background
There are several flood risk assessments available and the choice of assessment depends on your risk management and planning needs. The most widely used floods maps for the U.S. are those provided by the Federal Emergency Management Agency (FEMA), which utilizes historical local information for individual rivers and streams to model the flood risk. The information is then put together to form a nationwide map and is used for the National Flood Insurance Program (NFIP). To see if your home is in a flood risk zone, you can use the FEMA Flood Map Service Center (MSC) .
However there are other approaches that use large-scale, spatial information to provide details that the traditional approach to flood modeling may not be able to capture, such as the more recent assessments comducted by Wing et al. " Millions More Americans Face Flood Risks Than Previously Thought " and First Street . This is the reason why estimates for flood risks can be presented as a range, such as the 13 - 41 million mentioned previously.
However, my flood research has a different goal...
My research goal is not only to tell you WHERE there is flood risk, but to predict WHEN the area is at risk of flood.
Social Relevance
Floods are not just the result of hurricanes (as seen below) and occur more often than people realize. Flooding can occur even when there is not a rain drop in sight!
Due to the effects of climate change, rising global temperatures and sea-level rise, the occurrence of floods is expected to increase, especially for people living on coastal communities .
Floods have implications for health & safety as well as economic impacts. Over the past 30 years, an average of 86 people have died per year due to flood events and this number continues to increase.
Additionally, the economic costs of floods can (and do) exceed billions of dollars. Between 2016 and 2019, there were 9 billion dollar flooding events . Flooding impacts water supplies and economic activity, thus impacting quality of life.
"Water is our most precious resource and my mission is to provide solutions for sustainable and resilient water management." - Equisha Glenn
Research Question
Can modeling the spatial, simultaneous occurrence of floods improve seasonal flood predictions?
Traditionally, floods are modeled according to independent flooding events that occur at individual lakes and streams.
This research aims to show that there is a spatial relationship between floods that occur within a region. Understanding how the floods occur simultaneously and what links these concurring events can improve flood prediction and thus contribute to better water management practices.
This image is showing the location where flooding occurred at the same time within the Northeastern U.S.
Methods
To understand the characteristics of simultaneous, spatial floods this research uses the Northeastern region of the U.S., specifically the Hydrological Unit Code 2 (HUC2) region. Historical streamflow data was collected for the years 1955 – 2017 for 55 stations within the HUC2 region. Floods are classified as a station streamflow that exceeds the 99th percentile of flowrates that were measured during this period.
- To understand the spatial distribution of floods in the NE, we investigate the following:
- Where are simultaneous floods occurring?
- Are they randomly distributed or clustered?
- How often do they repeat? – frequency analysis on individual stations and jointly
- Peak
- Volume
- Duration
- Investigate local effects vs. synoptic and large-scale climate impacts
- Use historical and climate data to create flood risk prediction model
Results
Results showing how much of the Northeastern HUC2 area is in each risk category.
Initial results show that large spatial extent floods, that we have designated as Spatial Extreme Flood Events (SEFEs), occur as simultaneous floods of more than 3 stations at the same time. Binomial cumulative distribution analysis showed that the exceedance of 3 stations experiencing a flood on the same day is not the result of chance.
This resulted in 1,935 SEFEs occurring primarily during the season of January – May (JFMAM). Using correlation analysis, we found that El Nino Southern Oscillation (ENSO), Pacific Decadal Oscillation (PDO), and northeastern Caribbean sea-surface temperatures (SSTs) have a statistically significant correlation up to 1 season lag with SEFEs.
This means we can use this information to predict extreme flood conditions 6 months ahead of time. Therefore in the winter, we can provide flood risk conditions for the upcoming spring.
Future Work
Next steps include using pre-season climate predictors to forecast the number of large scale simultaneous floods from January to May.
A preliminary flood risk prediction tool will be developed and tested to see how well it can predict floods during the spring for the Northeast using the historical river/stream flood data and climate information.
These forecasts can then be used for quantifying the spatial extent and risks to prepare for water management and emergency services ahead of the season.
NERTO
My NOAA Experiential Research & Training Opportunity (NERTO) took place at NOAA's Center for Weather and Climate Prediction (NCWCP) located in College Park, MD. I had the oppornuity to spend 3 months at NCWCP and work with my NOAA mentor, Dr. Tom Smith, in addition to interacting and collaborating with NOAA scientists and forecasters.
NOAA Center for Weather and Climate Prediction located in College Park, MD | Source: NOAA
During my internship at NCWCP, my research focused on understanding the interactions between warming Caribbean sea-surface temperatures (SST), rainfall, and atmospheric conditions.
The Caribbean has been experiencing a decrease in rainfall over the past few years and my research aimed to understand the cause of this decrease. I used NOAA datasets and indices including optimum interpolated sea-surface temperature (SST) and the Galvez-Davison Index (GDI) for Tropical Convection.
Fun Fact...
I had the opportunity to collaborate directly with all of the scientists and forecasters who developed the the SST data and GDI!