Environmental Time Series Analysis

A Shiny web application by Sensing Clues for exploring and visualizing environmental remote sensing time series data, including NDVI (vegetation) and Burned Area analyses across multiple conservation project areas.

Features

NDVI Explorer

• Time Series — Plot multi-year NDVI trends for a selected project area and spatial resolution.
• Land Cover Explorer — Visualize NDVI trends broken down by land cover type (crops, rangeland, water, trees, flooded vegetation, built area, bare ground), alongside an interactive land use/land cover map.
• Delta Map — Generate static and interactive maps showing NDVI change (delta) for a selected month and year.

Burned Area Explorer

• Time Series — Plot burned area trends over time for a selected project area.
• Map Explorer — Generate static burned area maps and interactive maps for a specific month and year, with GeoJSON export functionality.

General

• Interactive map showing the selected Area of Interest (AoI).
• Figures are cached after first generation to speed up repeated requests.
• Multiple satellite data sources and spatial resolutions supported.

Supported Project Areas

Area	Country	Available in
Mponda	Zambia	NDVI Explorer
Ancares Courel	Spain	NDVI Explorer
Stara Planina	Bulgaria	NDVI Explorer
Kasigau	Kenya	NDVI Explorer
West Lunga	Zambia	Burned Area Explorer

Data Sources & Resolutions

Source	Resolutions
ESA Sentinel-2	100 m, 1000 m
Terra MODIS	250 m, 500 m, 1000 m

Data is available from 2015 onwards.

Project Structure

app/
├── app.R                  # Entry point
├── global.R               # Library loading and global variables
├── server.R               # Shiny server logic
├── ui.R                   # Top-level UI definition
├── ui/
│   ├── mod_header_ui.R    # Header module
│   ├── mod_sidebar_ui.R   # Sidebar controls module
│   └── mod_body_ui.R      # Main body/tab content module
├── src/
│   ├── utilities.R        # Helper functions
│   ├── visualization.R    # Visualization utilities
│   └── generate_plots.R   # Plot generation functions
├── translations.json      # UI translation strings (EN/NL/FR)
├── DESCRIPTION            # R package dependencies (deprecated)
└── www/
    ├── figures/           # Cached generated figures
    └── style.css          # Custom app styling

Installation & Running

Prerequisites

• R (>= 4.0 recommended)
• The following R packages (as loaded in global.R):

install.packages(c(
  "dplyr", "future", "ggplot2", "ipc", "jsonlite",
  "leaflet", "leaflet.extras", "lubridate", "promises",
  "raster", "shiny", "shinybusy", "shiny.i18n", "shinyjs",
  "shinyTree", "shinyWidgets", "sf", "terra", "tidyr",
  "plotly", "htmlwidgets", "devtools"
))

• The sensingcluesr package (installed from GitHub):

devtools::install_github("sensingclues/sensingcluesr@v1.0.3")

Run the App

# From the app/ directory
shiny::runApp("app")

Or open app.R in RStudio and click Run App.

Data Setup

The app reads input data from a folder with the following structure:

<data_dir>/
├── AoI/
│   └── AoI_<project_area>.geojson  # Area of Interest boundaries
├── BurnedArea/
│   └── <project_area>/
│       └── <resolution>/           # Burned area raster data files
├── LandUse/
│   └── <project_area>/
│       └── S2_10m_LULC_2023/       # Land use/land cover GeoJSONs
└── NDVI/
    └── <project_area>/
        └── <resolution>/           # NDVI raster data files

When running locally, you can place this data folder anywhere on your machine. Update the data_dir variable in global.R to point to it:

# global.R
data_dir <- file.path("path/to/your/local/data/folder")

A commented-out example is already provided in global.R for reference:

# Set test folder structure (uncomment when working locally with a different folder structure)
# test_dir <- file.path("www/data")
# data_dir <- test_dir

Generated figures are stored automatically in app/www/figures/.

Contact

For data issues or support, contact helpdesk@sensingclues.org.

More information: sensingclues.org/environmental-time-series-analysis