Description

Land cover and land use change (LCLUC) describes the many ways our world changes around us. Every new building, each vacant lot, and the expansion and retreat of forests or wetlands are all examples of LCLUC. These changes may be slow or sudden, but understanding the ways in which our landscape is changing gives us the opportunity to better anticipate and avoid problems that may arise. 

This training, led by NASA's Applied Remote Sensing Training program (ARSET), covers some of the ways that we can use NASA data to determine what is happening on the ground and how that changes over time. Participants will see examples of supervised and unsupervised classification models implemented in the R statistical coding language, and improve their understanding of how these models can be used to produce maps and metrics of LCLUC.

Prerequisites

• Fundamentals of Remote Sensing

Objectives

By the end of this training, attendees will be able to:

• Access NASA Earth observation data (e.g., Harmonized Landsat and Sentinel-2 (HLS)) relevant to LCLUC mapping.
• Convert NASA Earth observation data into distinct land cover and land use (LCLU) classes using supervised and unsupervised machine learning classification methods in the R programming language.
• Recognize the role of classification methods as one part of a change monitoring strategy.
• Compute a change matrix representing the change in LCLU between two dates.
• Create a map in RStudio visualizing the differences in LCLU between two dates.

Target Audience

• Professionals interested in coding to quantify land cover change as it relates to phenomena such as forest composition change, deforestation, urban expansion, habitat loss, and hydro/cryosphere changes.
• Other scientists or analysts interested in leveraging automated tools for decision support around land use management, natural resource management, or other issues that have an impact on how our world is structured across space and time.

Course Format

• Two 2-hour parts
• Two identical sessions will be hosted at two different times of day:
  • Session A: 11:00 a.m. to 1:00 p.m. EST (UTC-5)
  • Session B: 2:00 p.m. to 4:00 p.m. EST (UTC-5)
• Each part will include a 30-minute live Q&A.
• Those who attend all parts and complete the homework by the due date will receive a certificate of attendance.

Sessions

Part 1: Classification Methods for Land Cover

Tuesday, Feb. 24, 2026

• Why LCLU Matters
• Drivers of LCLU Change
• LCLUC Examples
• Spectral Profiles and Training Data
• Supervised and Unsupervised Classification
• Overview of Selected Classification Methods
• Accessing HLS Imagery with Earthdata Search
• Building a K-Means Clustering Model
• Interpreting Model Results
• Building a K-Nearest Neighbor Model
• Interpreting Model Results

ARSET Instructor: Justin Fain

Part 2: Visualizing Land Cover Change

Thursday, Feb. 26, 2026

• Review of Selected Classification Methods
• How Random Forest Models Work
• Building a Random Forest Model 
• Interpreting Model Results
• Applying Models Across Dates
• Creating a Change Matrix
• Mapping Changes in LCLU

ARSET Instructor: Justin Fain