This repository contains a project focused on evaluating plant phenology near the Santa Clara River, which spans from Santa Clarita to Ventura. Phenology, or the study of life cycle events, plays a crucial role in understanding ecological synchronization and its disruptions due to climate change. This project utilizes a time series of Landsat imagery to assess phenological patterns across diverse plant communities, with implications for broader ecological dynamics.
The project aims to analyze the phenological behaviors of distinct plant communities along the Santa Clara River by examining:
- Riparian forests characterized by winter deciduous species like cottonwood and willow trees
- Grasslands, primarily comprising drought deciduous grasses
- Chaparral shrublands, dominated by evergreen shrubs
├── data
│ ├── landsat-data
│ └── study_sites
├── analysis.RMD
├── analysis.html
├── .RData
├── .Rhistory
├── README.md
└── .gitignore
The analysis leverages the Landsat OLI sensor's capabilities, utilizing eight pre-processed scenes:
- Level 2 surface reflectance products with erroneous values set to NA
- Scaled reflectance values to 100 for consistency
- Spectral bands 2-7 used for detailed vegetation assessment
- Scene dates encoded in filenames for temporal tracking
- Study sites are mapped as polygons, each annotated with the plant community type to facilitate targeted analysis Data are stored in the .gitignore and can be downloaded through this link: https://drive.google.com/drive/folders/1ON8FbDqcTjg2PKHmNGgyN7odTqpOnXla
Landsat data: “Operational Land Imager.” NASA, December 9, 2021. https://landsat.gsfc.nasa.gov/satellites/landsat-8/spacecraft-instruments/operational-land-imager/.
The methodology involves:
- Transforming spectral reflectance into NDVI, a proxy for vegetation productivity
- Tracking NDVI variations over the year to capture phenological shifts
- Summarizing and comparing NDVI metrics within plant communities
- Visualizing NDVI trajectories to infer ecological responses to climate dynamics
R is the coding language used, along with libraries to handle geospatial data, visualization, and statistical analysis. The process encompasses:
- Defining a function to calculate NDVI from spectral bands
- Applying this function to individual scenes and generating NDVI layers
- Streamlining the NDVI computation across all scenes using a custom function
- Extracting NDVI values at specific study sites representing various vegetation types
- Cleaning and structuring the resultant data for clarity and ease of analysis
- Creating visualizations that show seasonal vegetation patterns across different ecosystems