Using Landsat Images, Postdoctoral Fellow Maps Damage from Gypsy Moth Infestation

Moth infestation map

Images and maps showing the magnitude and extent of damage from the current gypsy moth outbreak in southern New England created by Valerie Pasquarella, a postdoctoral fellow with the University of Massachusetts Amherst’s Northeast Climate Science Center, were featured as the U.S. Geological Survey “Image of the Week” during the week of Aug. 24.

Using time series of Landsat images, Pasquarella generates predicted images based on models fit to historical Landsat observations. These predicted images are then compared with actual images acquired on a given date, creating a series of difference maps that quantify potential gypsy moth damage in near-real-time. Individual near-real-time assessments can also be combined to produce season-integrated damage estimates like those shown as the Image of the Week.

She explains, “In 2016, we were piloting the time-series approach in response to an emerging outbreak. This year we focused on monitoring how the outbreak has evolved, and we’ve been shocked by how much more gypsy moth damage we’re seeing this summer.” 

She adds that the Image of the Week and accompanying video show both the 2016 and 2017 potential damage maps, and areas in red, orange and yellow, which correspond to different magnitudes of estimated damage, have spread considerably. “There was hope that the wet spring this year would revive the fungal biocontrol that had been keeping gypsy moth populations under control since the late 1980s, but fungal mortality didn’t peak until after the caterpillars had already caused a significant amount of damage,” she notes.

Further, Pasquarella says that “even with high mortality, with such a large population, we will likely see another gypsy moth defoliation event next year, though the magnitude and extent of damage remain uncertain.”

Pasquarella’s method is an improvement over traditional monitoring with aerial surveys that take hundreds of hours by pilots in small planes criss-crossing the state once per season. Her new method is based on Landsat images acquired every eight days and automated algorithms for extracting change information. She says that near-real-time monitoring and assessment of gypsy moth impacts provides valuable information not only for scientists but for the public about the extents and magnitude of ongoing outbreaks.

Link