Los Angeles: 37-year NDVI, LST & Vegetative Cooling (1985–2021)

Question: How have neighborhood greenness (NDVI), land-surface temperature (LST), and plant-driven cooling changed across Greater Los Angeles—and what’s driving those changes?

What we did: Built a multi-decadal Landsat time series (1985–2021) at census-tract scale, linked to weather (TerraClimate), drought (SPEI), land cover (NLCD), elevation (SRTM), and socio-demographics to quantify trends and drought responses

Outcome: L.A. warmed ~0.13 °C/yr, NDVI increased, and vegetative cooling strengthened, with equity-relevant patterns and a coast-to-inland drought gradient that can guide investments (e.g., canopy and heat mitigation).

A pair of maps of the greater Los Angeles urban region overlaying a background terrain. The left map shows NDVI (vegetation health) with green shades, and the right map shows LST (land surface temperature) with red and blue shades.

Research Challenge

Los Angeles is semi-arid, highly irrigated, and socio-demographically diverse. Decision-makers needed to understand where and why heat and greenness have shifted through time, including what happens during drought. The challenge was separating long-term climate and land-cover effects from year-to-year weather and surfacing actionable patterns at the tract scale across ~3,474 km² of urbanized area.

Approach

Data

  • Landsat ARD (4/5/7/8) for NDVI and LST (1985–2021).

  • TerraClimate (precipitation, temperature, radiation, etc.).

  • SPEIbase drought index (1–48-month scales).

  • NLCD (impervious %, tree canopy).

  • SRTM elevation; U.S. Census (income, race/ethnicity, education).

Methods

  • Cloud-filtered Landsat stack (215 images), tract-level aggregation, monthly series.

  • Trend analysis for NDVI, LST, and LST-NDVI slope (vegetative cooling).

  • Drought contrasts (wet vs. dry at SPEI-6), and regression/SEM to explain spatial variability (income, impervious, coast distance, canopy).

Tools

  • MATLAB (pre-processing, regression), ArcGIS Pro (spatial ops), Google Earth Engine (climate/elevation retrieval).

Maps and scatter plots showing the relationship between vegetation health, temperature, and climate variables in a geographic region. The top left map illustrates NDVI values with a color scale, while the bottom left map displays land surface temperature. The right side features three scatter plots showing correlations between NDVI, spring LST, and SPEI-6 over time.

Key Findings

  • Citywide trends (1985–2021).

    • LST +0.13 °C/yr; NDVI +5.05×10⁻⁴/yr.

    • Vegetative cooling (|LST–NDVI slope|) strengthened ~0.08 °C per year; net +2.96 °C/NDVI of cooling over the record.

    • Weather matters: NDVI tracks 3-mo precipitation; solar radiation + air temperature explain 87% of LST variability.

  • During drought (SPEI-6).

    • NDVI −0.023; LST +4.41 °C on average; vegetative cooling +0.08 °C/NDVI.

    • Strong coast-to-inland gradient: ~+8 °C inland vs +2–3 °C near the coast.

  • Drivers & equity signals.

    • Income, impervious cover, and distance from coast explain spatial variability in warming and greening trends; more greening → slower warming.

    • Long-term “luxury effect” weakened (income’s cooling: −41%; greening: −28% from 1990→2020).

Cover of a journal titled "Urban Climate" with a stylized silhouette of a city skyline in yellow and gray against a blue and purple sky background. The Elsevier logo is in the top left corner, and the ISSN number is in the top right corner. The editors-in-chief and their affiliations are listed at the bottom.
Two side-by-side box plots showing income disparities among different racial groups from 2000 to 2020. The left graph depicts income measured in NDVI, while the right graph shows income in Celsius. The racial groups include White, Hispanic, Black, and Asian, indicated by different colors.

Outputs

  • Deliverables: tract-level trend maps (NDVI, LST, cooling), drought response maps, equity overlays, methods brief, and reproducible analysis code.

  • Publication: Urban Climate (2023), “Urban greenness and its cooling effects are influenced by changes in drought, physiography, and socio-demographics in Los Angeles, CA.” DOI: 10.1016/j.uclim.2023.101743.

Impact / How it’s used

  • Provides a defensible baseline for urban heat and greenness trends citywide, with tract-level signals to target canopy and cooling investments, especially inland.

  • The drought gradient (coast vs. inland) and weakening luxury effect offer policy-relevant guidance for equity-minded planting/water strategies.

  • Methods underpin ongoing stakeholder work (e.g., tree-priority indices, urban fire interactions) and can be re-run for any city or AOI.