Is deforestation happening in this area, and what are its effects?

What you can answer

Where forest cover dropped in a specific area over the last N years (NDVI time-series, Hansen annual loss).
How much biomass was lost in tons of carbon (GEDI L4A footprint biomass averaged inside loss polygons).
Whether replant / regrowth has occurred (NDVI recovery curve post-event).
Downstream effects you can begin to test: surface temperature change (ECOSTRESS), evapotranspiration change (ECOSTRESS), soil moisture change (SMAP), precipitation runoff change (IMERG climatology shift).

What you can NOT answer with these datasets alone

Direct cause (logging vs fire vs disease vs urban expansion) — needs ancillary data.
Sub-pixel canopy change smaller than ~30m (HLS resolution floor).
Continuous biomass loss in time — GEDI is footprint-sampled, not continuous swath.

Code template (Python, cloud-direct, ~30 lines)

import earthaccess
import xarray as xr
from pystac_client import Client

earthaccess.login(strategy="netrc")

# 1. Define AOI + time window
aoi = (78.5, 15.0, 73.5, 20.0)  # W, S, E, N — Maharashtra Western Ghats
years = (2018, 2024)

# 2. Pull HLS L30 (Landsat) surface reflectance for NDVI
hls_results = earthaccess.search_data(
    short_name="HLSL30",
    bounding_box=aoi,
    temporal=(f"{years[0]}-01-01", f"{years[1]}-12-31"),
    cloud_cover=20,
)
# ...open into xarray, compute NDVI = (NIR - Red) / (NIR + Red), tile-mosaic per year

# 3. GEDI L4A footprint biomass intersecting AOI
gedi = earthaccess.search_data(
    short_name="GEDI_L4A_AGB_Density_V2_1_2056",
    bounding_box=aoi,
    temporal=(f"{years[0]}-01-01", f"{years[1]}-12-31"),
)
# ...open HDF5, extract `agbd` and `agbd_se` at footprint coords

# 4. Mask GEDI footprints by HLS-derived loss polygons → compute mean biomass lost
# 5. Plot: NDVI time-series + biomass-density-lost histogram

Expected output

Map: change in mean annual NDVI between (years[0], years[0]+2) baseline and (years[1]-2, years[1]) recent
Histogram: biomass density (Mg/ha) of GEDI footprints inside detected loss polygons
Time-series: pixel-mean NDVI over the AOI, with annotation at known loss events

Caveats

HLS cloud-cover filter at 20% is aggressive; cloudy regions (Western Ghats monsoon) lose most of June–Sept observations.
GEDI ended primary mission in March 2023; data from 2019–2023 only for the original era (extended mission resumed 2024 onboard ISS).
Hansen Global Forest Change is annual, defines “loss” as ≥50% canopy reduction in a 30m pixel — coarser than typical small-holder agroforestry change.

Cross-DAAC composition

This is a 3-DAAC join: LP DAAC (HLS, MODIS) + ORNL DAAC (GEDI). Auth is uniform (Earthdata Login) but the access libraries differ (earthaccess for both, but HDF5 reading for GEDI is more involved than COG-mosaic for HLS). See recipes/r01-three-daac-composition.mdx for the general pattern.