Last updated: January 20, 2025
Article
Seasonal Variations in Remotely Sensed Data
The data showed more buildings and fewer trees in one season than in another—but not because of land use changes.
By the editors of Park Science magazine
Image credit: NPS / Kelsey Graczyk
Remote sensing data—images of the earth gathered by satellite or aircraft—can show us what’s on the land and how people use the land. They show, for example, whether an area has forest, urban development, or farmland. This information can be helpful in studying things like water quality and climate change. But using data from different seasons can cause mistakes when classifying different land types. This has real-world consequences.
Scientists discovered that remote sensing data from the same area resulted in different land use/land cover (LULC) classifications depending on season. The data reflected more buildings and fewer trees in one season than in another season. But this wasn’t because the land use had actually changed. They describe their work in the December 2024 issue of the journal, Hydrology and Earth System Sciences.
Researchers often use mathematical models that incorporate LULC data to analyze or predict real-world outcomes for different environmental scenarios. Land managers use those results to make decisions about natural resources. But models that use LULC data from different seasons could yield erroneous results, leading to less robust decisions. The study authors give tips on how to get more accurate results from LULC data obtained during different seasons.
Myers and others. 2024. Seasonal variation in land cover estimates reveals sensitivities and opportunities for environmental models. Hydrology and Earth System Sciences 28: 5295–5310.