Objective: The objective of the project was to use concepts in remote sensing to record, detect, and analyze changes in land use patterns over time. Two multispectral satellite images of northern Costa Rica were compared from 2022 and 2001 to capture largescale agricultural expansion in the country. Remote sensing was crucial to this analysis as it was able to show specifically how and where land use change has been occurring in Costa Rica.

Methodology: Multispectral images were converted from digital number values (DN) to reflectance values. A supervised image classification was then run to assign all pixels in the image to a given landscape feature class. A change detection analysis measuring the difference in area for each class revealed land use change in Costa Rica has been dominated by agricultural consolidation. Although it was expected to see an overall increase in agricultural land and decrease in forests over time, the analysis showed that this was not the case. Instead, smaller agricultural plots that had already existed in the past were consolidated to build larger monocropping farms.

Created for: Remote sensing course final project

Software and tools used: ERDAS Imagine- supervised/unsupervised classification, index analysis, DN to reflectance

Credit: Matt Manfredi was a group partner on this project