Coastal Change Monitoring with NDWI and SAR Time Series

The Urgency of Coastal Monitoring

Coastlines are simultaneously some of the most densely populated and most threatened environments on Earth. Sea level rise, storm surge intensification, erosion, and human modification are reshaping coastal zones faster than traditional survey methods can document.

Global studies estimate that between 24% and 33% of sandy beaches are eroding. Mangrove loss exceeds 1% per year in some regions. Coral reef extent declines in response to thermal bleaching events that occur on 5–7 year cycles.

Satellite time series monitoring cannot stop these changes, but it can document them objectively, measure rates precisely, and provide the spatial detail needed to prioritize conservation and adaptation investments.

Key Coastal Monitoring Applications

Shoreline Change Tracking

The water-land boundary moves naturally with tides and seasonally with sea level, but long-term shoreline position changes reflect erosion, accretion, or sea level rise.

Monitoring approach:

• Use MNDWI over a polygon spanning the shoreline and a buffer on each side
• Monthly or seasonal averages smooth tidal variation
• Long-term trends (5+ years) reveal directional change

A declining MNDWI average (more land visible in the water-land zone) indicates accretion. An increasing MNDWI average indicates shoreline retreat or erosion.

SAR complement: SAR VV does not require cloud-free conditions and provides consistent observations regardless of tidal timing. Consistent observation timing removes tidal contamination that affects optical MNDWI data.

Storm Surge and Coastal Flooding

Storm surge flooding is temporary but can be catastrophic. Monitoring the transition:

Before storm: Normal MNDWI baseline for the coastal zone During surge: Sharp MNDWI increase as seawater floods normally dry land After recession: MNDWI return to baseline; SAR VV recovery

SAR advantage during storms: Storms bring clouds that block optical satellite observation exactly when flooding is occurring. SAR provides observation even during cloudy storm conditions, giving flood extent measurements that optical sensors cannot.

Cross Ratio (CR = VH/VV): In flooded vegetation (mangroves, coastal marshes), the SAR signal changes in a distinctive way — double-bounce scattering from the water surface beneath the canopy increases VH relative to VV, raising CR. Open water has low CR; flooded vegetation has elevated CR. This distinction is important for accurately mapping inundation extent in vegetated coastal areas.

Mangrove Extent and Health

Mangroves are among the most carbon-dense ecosystems on Earth and provide critical coastal protection. But they are also among the most threatened — by aquaculture conversion, urban expansion, and climate stress.

Monitoring mangroves with combined indices:

MNDWI: Mangrove zone is water-land boundary. As mangroves are cleared, MNDWI increases (more open water or mud exposed).

NDVI: Healthy mangroves have high NDVI (0.5–0.8). Dying or degraded mangroves show declining NDVI before total loss.

SAR RVI: Mangrove volume scattering produces high RVI. Clearing reduces RVI immediately and permanently.

SAR RFDI: Partial degradation (disease, selective cutting) shows elevated RFDI before complete clearing is visible.

By tracking NDVI, RVI, and RFDI simultaneously, you can detect mangrove degradation earlier than any single index alone.

Coral Reef Zone Monitoring

Direct coral reef monitoring from satellite is constrained by water depth attenuation and resolution, but satellite indices can monitor:

Turbidity: Sediment runoff following rainfall or coastal disturbance increases water turbidity, detected as reduced apparent water-leaving reflectance

Algal blooms: Excessive nutrients stimulate surface algal growth visible as green patches in NDWI data

Bleaching signatures: Thermal bleaching is best tracked via sea surface temperature data, but post-bleaching loss of coral tissue and algal overgrowth produces changes in reflectance detectable with Sentinel-2

Tidal Flat and Intertidal Zone

Tidal flats are exposed at low tide and submerged at high tide. Monitoring MNDWI over a tidal flat zone averages over multiple tidal states, providing:

• Mapping of the mean intertidal zone extent
• Detection of changes from sediment accretion or erosion
• Tracking of reclamation activities (land fill of tidal flat for development)

Over years, declining tidal flat extent in MNDWI data may indicate:

• Natural sediment dynamics
• Human land reclamation
• Sea level rise reducing the intertidal exposure window

Setting Up Coastal Monitoring

Polygon Design

For shoreline tracking: Draw the polygon to span the shoreline position, extending ~500m to 1km inland and into the water. This ensures both land and water are captured within the polygon, so the average MNDWI value reflects the water-land balance.

For storm surge areas: Draw over the low-lying coastal zone likely to flood during storm events. Include areas up to 2–3m elevation above current mean sea level.

For mangroves: Draw a polygon closely following the mangrove fringe. Too large a polygon dilutes the mangrove signal with open water or upland vegetation.

Index Selection for Coastal Work

A comprehensive coastal monitoring setup includes:

• MNDWI (primary water extent indicator)
• NDWI (comparison/backup)
• NDVI (vegetation health including mangroves)
• Sentinel-1 VV (all-weather surface change)
• Sentinel-1 CR (flooded vegetation detection)

Start Date for Coastal Monitoring

• Short-term event response (storm damage, oil spill): Start 3–6 months before the event
• Seasonal dynamics (tidal flat changes, mangrove growth): 12–24 months back
• Long-term erosion/accretion trends: 3–5 years back

Limitations

Tidal contamination in optical data: Tidal state at the time of satellite overpass affects apparent MNDWI values. Ideally, compare same-tidal-state observations. In practice, long-term averages smooth this effect.

Turbidity variation: Turbid water has different spectral properties than clear water, affecting MNDWI accuracy. Heavy sediment loads can suppress apparent MNDWI.

Resolution for fine-scale features: At 10m (Sentinel-2) or 5×20m (Sentinel-1 IW mode), narrow mangrove fringes or small tidal channels may not be resolved. This is suitable for landscape-level monitoring, not individual tree assessment.

Saltwater vs. freshwater: NDWI and MNDWI respond similarly to both; they cannot distinguish salinity from extent. Freshwater coastal wetlands and saltwater coastal zones look similar spectrally.

Summary

Coastal change monitoring with MNDWI and Sentinel-1 SAR time series provides all-weather, continuous tracking of one of Earth's most dynamic and threatened environments. MNDWI tracks water extent changes including shoreline position and flood extent; SAR VV provides cloud-independent observations; SAR CR distinguishes flooded vegetation from open water; NDVI and RVI track mangrove canopy health. The combination provides a multi-dimensional picture of coastal change that no single sensor can deliver alone.