Abstract
Satellite imagery is a practical and valuable tool for monitoring vegetation condition in forests. The longevity of the USGS/NASA Landsat program along with its medium spatial resolution (30m) gives researchers the ability to make informed statements on land cover generally, and specifically on aspects such as forest conditions. The Landsat program’s nearly 50-year archive of imagery show how Earth’s surface has changed through modern development and how these developments have influenced forests. Google Earth Engine (GEE) is a cloud-based repository of satellite imagery dating as far back as the 1970s. This study utilizes Landsat 5-8 imagery from GEE to calculate the long-term vegetation structure trends in Berea College Forest (BCF) in Berea, Kentucky from 1984-2020. By calculating the average growing-season Normalized Difference Vegetation Index (NDVI) and using the Mann-Kendall trend test and Sen’s slope estimator, I evaluated the significance of vegetation productivity trends on a pixel-by-pixel basis. The results show that 68.47% of BCF displayed significant trends in NDVI, with most of these pixels associated with a positive trend, and NDVI values for the study area increased at a rate of 0.001985 units per year. These positive trends were mostly clustered in the northern head and eastern tail of BCF. The southern portion displayed a clustering of pixels with no significant trend. Significant negative trends were rare but present. The most noticeable negative trend is attributed to US Highway 421, which began construction in 1998. Understanding long-term vegetation dynamics in BCF will assist foresters in developing effective management plans.
Recommended Citation
Foushee, Jacob
(2021)
"Multi-Decadal Analysis of Remotely Sensed Vegetation Change in Berea College Forest - Seasonality of Forest Patterns using Remote Sensing,"
The Cardinal Edge: Vol. 1:
Iss.
2, Article 18.
Available at:
https://ir.library.louisville.edu/tce/vol1/iss2/18