Executive Summary · May 2026

Field Guide to Copernicus
Remote Sensing and Light Indices

Revised Edition. Practical, field-usable guidance for advanced students, community scientists, and watershed practitioners.

Purpose

Provide a defensible, field-manual workflow for bloom screening, water masking, and cyanobacteria inference using Copernicus imagery. Emphasis: separate water detection, chlorophyll screening, and phycocyanin logic.

What Changed

Expanded index catalog, added CDOM blue-loss page, added direct phycocyanin methods (Sentinel-3 only), and separated proxy vs direct indicators. Added executive report tab and troubleshooting page.

Bottom Line

Color alone is not proof of a bloom. Sentinel-2 can screen for biomass but cannot directly measure phycocyanin. Sentinel-3 enables stronger phycocyanin logic at 300 m resolution.

Operational stack (rivers)

Step	Primary Index	Goal	Notes
1	MNDWI → NDWI	Water mask	Mask emergent vegetation and shoreline mixing
2	NDCI	Bloom screen	Chlorophyll proxy; not phycocyanin
3	FAI / SABI	Surface scum	Detect floating mats
4	CCI / S2-MPH / S2-FLH	Cyano hint	Heuristic only on S2
5	NDTI	Turbidity check	Filter sediment-driven false positives

Critical Caveat

Sentinel-2 MSI does not directly measure phycocyanin. All Sentinel-2 cyanobacteria mapping is indirect and proxy-based. Sentinel-3 OLCI supports stronger phycocyanin-oriented approaches using 620 nm and 709 nm bands. Do not overstate certainty; always field-verify.

Revised Edition · May 2026

Rapid Cheat Sheet

One-page reference for Copernicus water, bloom, and cyanobacteria screening.

Sentinel-2 Band Cheat Sheet

B01
443nm
60m
Aerosol

B02
490nm
10m
Blue

B03
560nm
10m
Green

B04
665nm
10m
Red

B05
705nm
20m
RedEdge1

B06
740nm
20m
RedEdge2

B07
783nm
20m
RedEdge3

B08
842nm
10m
NIR

B8A
865nm
20m
NarrowNIR

B09
945nm
60m
H2O

B10
1375nm
60m
Cirrus

B11
1610nm
20m
SWIR1

B12
2190nm
20m
SWIR2

Quick Index Table

Index	Purpose	Formula	Sensor	Validation
NDWI	Water mask	(G-N)/(G+N)	S2 / 10m	Established
MNDWI	Water mask (urban)	(G-S1)/(G+S1)	S2 / 20m	Established
WIW	Wetland water	NIR<0.1804 AND SWIR2<0.1131	S2 / 20m	Established
NDTI	Turbidity	(R-G)/(R+G)	S2 / 10m	Adapted
NDCI	Bloom screen	(RE1-R)/(RE1+R)	S2 / 20m	Validated
FAI	Floating scum	N - baseline(R,SWIR)	S2 / 20m	Established
PCI	PC proxy	24.665*(B5/B4)^3.4607	S2 / 20m	Validated
CI-cyano	Cyano index	-SS(681)	S3 / 300m	Validated
Simis05	Direct PC	a_PC(620) model	S3 / 300m	Direct (S3)
S2-MPH	Cyano peak shape	max(B05-07)-baseline	S2 / 20m	Adapted

Aquatic Confounders

What You See	Potential False Positive	How to Distinguish
High NDCI in dark water	CDOM or shadow	Check blue loss, NIR suppression
Bright green surface	Duckweed or green algae	Use CCI, S2-MPH, FAI
Tan plume	Suspended sediment	NDTI and true color
Bright shoreline pixels	Emergent vegetation	MNDWI water mask
Silver patches	Sunglint or foam	Discard scene

Critical Caveat

Sentinel-2 does not directly measure phycocyanin. All S2 cyanobacteria mapping is indirect and proxy-based. Sentinel-3 OLCI supports stronger phycocyanin logic using 620 nm and 709 nm bands. Do not overstate certainty.

Chapter 1

Core Concepts

Light as data, water optics, pigment absorption, and the red-edge logic that powers bloom screening.

Light as Environmental Data

Photon data

Remote sensing reads how surfaces absorb and reflect light. Each wavelength carries information about pigments, water, and particles.

Reflectance

Fraction of incoming light reflected. Healthy vegetation reflects NIR strongly and absorbs red.

Absorption

Chlorophyll-a absorbs ~440 nm (blue) and ~680 nm (red). Phycocyanin absorbs ~615–621 nm.

Scattering

Turbidity elevates visible reflectance and can create bloom false positives. Always check NDTI.

Red edge

Rapid rise between red and NIR. Sentinel-2 B05–B07 capture this; it powers NDCI and S2-MPH.

CDOM effect

CDOM absorbs blue light first. Tea-colored water is often CDOM, not a bloom.

Growth phase matters

Phycobiliprotein peaks near 625–627 nm can weaken under nutrient stress or stationary phase. Optical signals are not static across bloom phases.

Chapter 2

CDOM and Blue-Loss

CDOM is the absence of blue. Treat tea-colored water as a separate optical regime.

What CDOM Does Optically

Absence of blue

CDOM strongly absorbs blue light first, then green. The result is blue loss and dark, tea-colored water.

Tea-colored optics

Yellow-brown water is typical of tannin-rich runoff and wetlands. It can be biologically clean.

Why dark water is not automatically a bloom

Dark water can be CDOM-rich, not cyanobacteria-rich. Blooms typically raise red-edge reflectance; CDOM suppresses visible bands. Always check NIR behavior and true color before calling a bloom.

How CDOM confounds bloom screening

CDOM suppresses visible-band reflectance and can reduce NDCI/CCI values. A bloom in CDOM-rich water can be underestimated or missed. Treat CDOM as a separate optical regime.

Field cue

Tea-colored water plus low NIR reflectance typically indicates CDOM, not cyanobacteria.

Chapter 3

Copernicus Basics

Sentinel-2 and Sentinel-3 capabilities, band relevance, and resolution limits for aquatic work.

Sentinel-2 MSI — Key Bands for Water and HABs

Band	Name	Center (nm)	Res	Use
B02	Blue	490	10m	Water color, CDOM
B03	Green	560	10m	NDWI, turbidity
B04	Red	665	10m	Chlorophyll absorption baseline
B05	Red Edge 1	705	20m	NDCI, PCI, S2-FLH
B06	Red Edge 2	740	20m	S2-MPH peak shape
B07	Red Edge 3	783	20m	S2-MPH peak shape
B08	NIR	842	10m	NDVI, FAI, water masking
B11	SWIR1	1610	20m	MNDWI, FAI, SABI
B12	SWIR2	2190	20m	WIW, water masking

Sentinel-3 OLCI — Aquatic Bands

Band	Center (nm)	Res	Use
Oa06	620	300m	Direct PC (Simis05), SY00 ratio
Oa07	665	300m	CI-cyano baseline
Oa08	674	300m	CI-cyano spectral shape
Oa11	709	300m	CI-cyano, Simis05, SY00

Direct PC logic

Sentinel-3 OLCI is the only Copernicus sensor with a 620 nm band for phycocyanin. Use Simis05 or SY00 ratio only on large water bodies.

Chapter 4

Light Indices Guide

Expanded index catalog with plain language, technical detail, formulas, band mapping, and validation flags.

Do not over-interpret

Only Sentinel-3 supports direct phycocyanin logic. Sentinel-2 indices are proxy or heuristic signals and must be field-verified.

Water / Masking / Aquatic Context

Index	Plain Language	Technical	Formula	Bands	Sensor	Res	Detects	Why It Works	Best Use	Confounders	Cautions	Category	Validation
NDWI	Basic open-water detector	Green vs NIR contrast	(G-N)/(G+N)	B03,B08	S2	10m	Open water	Water absorbs NIR	Initial water mask	Floating veg, shallow bottom	Use with MNDWI	water	Established
MNDWI	Water mask in urban/veg areas	Green vs SWIR	(G-S1)/(G+S1)	B03,B11	S2	20m	Open water	Water absorbs SWIR	River and lake masks	Wet soil, emergent veg	Threshold per scene	water	Established
AWEIsh	Water mask in shadow	Multi-band linear model	B+2.5G-1.5(N+S1)-0.25S2	B02,B03,B08,B11,B12	S2	20m	Water in shadowed scenes	Visible vs IR contrast	Complex terrain	Bright sediment	Not a bloom tool	water	Adapted
AWEInsh	Water mask (no shadow)	Multi-band linear model	-0.0128B+0.1406G+0.1762N-0.7879S1+0.1892S2	B02,B03,B08,B11,B12	S2	20m	Water in open terrain	Water absorbs NIR/SWIR	Flat watersheds	Bright soil	Not a bloom tool	water	Adapted
WIW	Water under vegetation	Threshold on NIR/SWIR2	NIR<0.1804 AND SWIR2<0.1131	B08,B12	S2	20m	Wetland water	Water absorbs NIR/SWIR	Wetland masking	Dark vegetation	Context only	water	Established
NDPI	Small pond detector	SWIR-green contrast	(S1-G)/(S1+G)	B11,B03	S2	20m	Small ponds, shallow pools	Water absorbs SWIR	Wetland pools	Wet soil	Use with NDWI	water	Adapted
S2WI	Water in vegetated riparian zones	Red-edge vs SWIR	(RE1-S1)/(RE1+S1)	B05,B11	S2	20m	Riparian water	Red-edge separates vegetation	Riparian masking	Emergent veg	Context only	water	Adapted
SWM	Composite water mask	Green + SWIR ratio	(G+S1)/(N*S2)	B03,B11,B08,B12	S2	20m	Water extent	Water absorbs NIR/SWIR	Automated pipelines	Bright wet soil	Threshold tuning	water	Adapted
LSWI	Surface wetness	NIR vs SWIR moisture ratio	(N-S1)/(N+S1)	B08,B11	S2	20m	Wetness and water content	SWIR sensitive to water	Seasonal wetness	Dense vegetation	Not bloom tool	water	Established
MBWI	Water boundary enhancer	Visible minus IR composite	2G-R-N-S1-S2	B03,B04,B08,B11,B12	S2	20m	Complex shorelines	Water bright in visible, dark in IR	Boundary mapping	Foam, sediment	Threshold tuning	water	Adapted
WRI	Simple water ratio	Visible vs NIR/SWIR ratio	(G+R)/(N+S1)	B03,B04,B08,B11	S2	20m	Water presence	Water is dark in IR	Quick check	Wet soil	Use with NDWI	water	Adapted
ANDWI	Multi-band water mask	Visible minus IR normalized	(B+G+R-N-S1-S2)/(B+G+R+N+S1+S2)	B02,B03,B04,B08,B11,B12	S2	20m	Water in mixed landscapes	Combines visible water brightness	Urban river masking	Bright sediment	Threshold tuning	water	Adapted
NWI	Shallow water detector	Blue vs IR normalized	(B-N-S1-S2)/(B+N+S1+S2)	B02,B08,B11,B12	S2	20m	Shallow/coastal water	Blue reflects, IR absorbed	Shallow boundaries	Bright sand	Depth-sensitive	water	Adapted
TWI	Triangle water index	Green normalized by RGB+NIR	G/(N+R+G)	B03,B08,B04	S2	10m	Water separation	Water dark in NIR	Narrow rivers	Floating vegetation	Context only	water	Adapted
FDI	Floating debris discriminator	NIR uplift above baseline	N - (RE3 + (S1-RE3)*(833-783)/(1610-783))	B08,B07,B11	S2	20m	Floating debris	Debris boosts NIR	Debris vs algae check	Floating veg	Use with FAI	floating	Adapted

Turbidity / Sediment Context

Index	Plain Language	Technical	Formula	Bands	Sensor	Res	Detects	Why It Works	Best Use	Confounders	Cautions	Category	Validation
NDTI	Turbidity proxy	Red-green ratio	(R-G)/(R+G)	B04,B03	S2	10m	Suspended sediment	Sediment elevates red	Storm runoff screening	Dense blooms	Flag false positives	turbidity	Adapted

Bloom / HAB / Cyanobacteria

Index	Plain Language	Technical	Formula	Bands	Sensor	Res	Detects	Why It Works	Best Use	Confounders	Cautions	Category	Validation
NDCI	Chlorophyll bloom screen	Red-edge vs red ratio	(RE1-R)/(RE1+R)	B05,B04	S2	20m	Chl-a proxy	Red-edge rises with chl-a	Primary bloom screen	CDOM, turbidity	Not phycocyanin	bloom	Validated
FAI	Floating scum	NIR uplift above baseline	N - [R + (S1-R)*(833-665)/(1610-665)]	B04,B08/B8A,B11	S2	20m	Surface mats	Floating biomass boosts NIR	Scum detection	Debris, floating veg	Misses subsurface blooms	floating	Established
PCI	PC proxy	Empirical power-law ratio	24.665*(B5/B4)^3.4607	B05,B04	S2	20m	PC proxy	PC affects red-edge ratio	Sturgeon Pool	CDOM, turbidity	Indirect on S2	cyano proxy	Validated
CCI	Red-green pigment hint	Red vs green ratio	(R-G)/(R+G)	B04,B03	S2	10m	PC dominance hint	PC shifts red-green balance	Post-screening check	Sediment, vegetation	Heuristic only	cyano proxy	Heuristic
PCINDEX	3-band PC proxy	Inverse red-edge difference	(1/RE1-1/RE2)*G	B05,B06,B03	S2	20m	PC proxy	Red-edge curvature vs green	Secondary PC check	Mixed pixels	Needs local calibration	cyano proxy	Adapted
SCBAI	Surface bloom anomaly	NDCI minus NDVI	NDCI - NDVI	B05,B04,B08	S2	20m	Surface cyano anomaly	Subtracts vegetation signal	Surface anomaly scan	Floating vegetation	Heuristic only	bloom	Heuristic
SNCI	Scum-weighted NDCI	NDCI multiplied by NIR	NDCI*N	B05,B04,B08	S2	20m	Surface scum severity	NIR rises with scum	Scum severity mapping	Floating vegetation	Heuristic only	floating	Heuristic
FCA	Composite severity score	Multi-index fusion	(NIR/Green)+NDVI+Blue+EVI+SABI+CCI	B02,B03,B04,B08,B11	S2	10m	Bloom severity proxy	Combines multiple signals	Dashboard visualization	Vegetation, turbidity	Composite proxy	bloom	Adapted
S2-MPH	Red-edge peak shape	Peak height above baseline	max(B05,B06,B07)-baseline(B04,B8A)	B04,B05,B06,B07,B8A	S2	20m	Cyano vs algae hint	Peak shifts to B06/B07	Type discrimination	Mixed pixels	Adapted tool	cyano proxy	Adapted
S2-FLH	Fluorescence suppression	Fluorescence line height	B05 - B04 - (B06-B04)*(705-665)/(740-665)	B04,B05,B06	S2	20m	Fluorescence behavior	Cyano can suppress FLH	Paired with NDCI	Low biomass, turbidity	Adapted tool	cyano proxy	Adapted
CI-cyano	EPA CyAN index	Negative spectral shape at 681 nm	-(Oa08 - Oa07 - (Oa11 - Oa07)*((681-665)/(709-665)))	Oa07,Oa08,Oa11	S3	300m	Cyano presence	Cyano suppresses fluorescence	Large lakes	Mixed pixels	Too coarse for rivers	cyano	Validated
MCI	S3 bloom delineation	Peak height near 709 nm	R10 - R09*(R11-R10)/(R11-R09)	S3 OLCI	S3	300m	Dense surface blooms	Red-edge peak height	Large lake mapping	Mixed pixels	Not for rivers	bloom	Established
Simis05	Direct PC retrieval	Semi-analytical absorption model	a_PC(620) = [R(709)/R(620)](a_w(709)+b_b) - b_b - a_w(620); PC = a_PC / a_PC	Oa06,Oa11	S3	300m	Phycocyanin absorption	Uses 620 nm PC band	Large lakes only	Mixed pixels, turbidity	Requires calibration	direct PC	Direct (S3)
SY00	PC ratio proxy	709/620 reflectance ratio	R(709)/R(620)	Oa11,Oa06	S3	300m	PC proxy	PC absorbs at 620 nm	Quick S3 check	CDOM, mixed pixels	Ratio only	cyano proxy	Adapted

Vegetation / Red-Edge Context (Confusion Removal)

Index	Plain Language	Technical	Formula	Bands	Sensor	Res	Detects	Why It Works	Best Use	Confounders	Cautions	Category	Validation
EVI2	Two-band canopy index	NDVI variant without blue	2.5*(N-R)/(N+2.4R+1)	B08,B04	S2	10m	Vegetation vigor	NIR-red contrast	Riparian masking	Bright soil	Not for water	vegetation	Established
GNDVI	Chlorophyll-sensitive	Green vs NIR ratio	(N-G)/(N+G)	B08,B03	S2	10m	Chlorophyll variation	Green reflectance shifts	Vegetation masking	Green algae on water	Use with water mask	vegetation	Established
NDRE	Red-edge chlorophyll	Red-edge slope metric	(RE2-RE1)/(RE2+RE1)	B07,B05	S2	20m	Canopy chlorophyll	Red-edge shifts with chl	Vegetation confusion removal	Mixed pixels	Not a bloom index	vegetation	Established
IRECI	Chlorophyll content proxy	Red-edge composite ratio	(RE3-R)/(RE1/RE2)	B07,B04,B05,B06	S2	20m	Canopy chlorophyll	Red-edge vs red absorption	Vegetation masking	Mixed pixels	Not a bloom index	vegetation	Adapted
S2REP	Red-edge position	Inflection point estimator	705+35(0.5(RE3+R)-RE1)/(RE2-RE1)	B05,B06,B07,B04	S2	20m	Vegetation stress	Red-edge shift indicates stress	Riparian vegetation context	Mixed pixels	Not a bloom index	vegetation	Established
ARVI	Haze-resistant NDVI	Blue-band aerosol correction	(N-(2R-B))/(N+(2R-B))	B08,B04,B02	S2	10m	Vegetation in haze	Blue corrects aerosols	Hazy scene masking	Water pixels	Use with water mask	vegetation	Established
SIPI	Pigment ratio proxy	Carotenoid vs chlorophyll	(N-B)/(N-R)	B08,B02,B04	S2	10m	Early vegetation stress	Pigment ratio changes early	Riparian stress	Water pixels	Not for water	vegetation	Established
PSRI	Senescence indicator	Red-green normalized by NIR	(R-G)/N	B04,B03,B08	S2	10m	Vegetation senescence	Carotenoid shift	Riparian context	Algae in water	Context only	vegetation	Established
MCARI	Chlorophyll absorption proxy	Red-edge correction index	((RE1-R)-0.2(RE1-G))(RE1/R)	B05,B04,B03	S2	20m	Canopy chlorophyll	Minimizes non-photosynthetic material	Vegetation masking	Mixed pixels	Not for water	vegetation	Adapted
MTCI	Red-edge chlorophyll proxy	MERIS-based red-edge ratio	(RE2-RE1)/(RE1-R)	B06,B05,B04	S2	20m	Canopy chlorophyll	Red-edge slope increases with chl	Vegetation context	Mixed pixels	Not for water	vegetation	Adapted

Chapter 5

Field Interpretation

Visual scenarios and field notes for rapid interpretation.

Bright Green Water Surface

True color: vivid lime or pea-soup green

Think: Dense phytoplankton or floating algae. High bloom risk.

Check NDCI first (bloom screen)
Use CCI and S2-MPH for cyano vs green algae hints
FAI elevated = surface scum
Do not assume toxicity

Dark Tea-Colored Water

True color: brown or amber

Think: CDOM and blue loss, not automatically a bloom.

Low NIR reflectance is typical
NDCI may be suppressed
Check CDOM context before flagging bloom

Tan Sediment Plume

Fan-shaped plume from tributary

Think: Suspended sediment, not bloom.

NDTI elevated
Runoff or dam releases likely
Mask before bloom analysis

Silver or White Patches

Bright, irregular highlights on water

Think: Sunglint or foam, not bloom.

Discard scene for bloom work
Check sun angle and wind

Chapter 6

If You See This, Think This

Fast pattern recognition for field teams.

If you see...	Think this...	Check with...
Bright green scum	Floating algae or cyano scum	FAI + NDCI + CCI
Tea-colored water	CDOM / blue loss	True color + low NIR
Tan plume	Suspended sediment	NDTI + true color
Green edges only	Emergent vegetation or mixed pixels	MNDWI water mask
Silver glare	Sunglint or foam	Discard scene
High NDVI over water	Duckweed or floating vegetation	FAI + CCI + NDWI

Chapter 7

Copernicus Browser Workflow

Step-by-step field workflow before sampling.

Step-by-Step

Weather & flow: Review gauges and recent storms.
True color: Identify CDOM, sediment plumes, sunglint.
Water mask: NDWI/MNDWI first.
Bloom screen: NDCI.
Surface scum: FAI or SABI.
Cyano hint: CCI or S2-MPH (heuristic).
Turbidity check: NDTI.
Document: Save screenshots + index values.
Field verify: Sample highest-signal pixels.

Chapter 8

Which Index Should I Use?

Decision guide for water, bloom, and cyanobacteria questions.

Decision Rules

Water mask first, bloom screen second, type discrimination third. Use Sentinel-3 only for large water bodies.

Water extent? NDWI → MNDWI → WIW (wetlands).

Bloom screening? NDCI, then FAI for surface scum.

Cyano vs green algae? S3: Simis05 or CI-cyano. S2: S2-MPH or CCI (heuristic).

Debris confusion? FDI + true color.

Turbidity spikes? NDTI before interpreting NDCI.

Chapter 9

Troubleshooting False Positives

Common error sources and corrections.

Trigger	Index Affected	Diagnosis	Fix
CDOM	NDCI, CCI	Dark water, low NIR	Treat as CDOM, do not over-interpret
Sediment plume	NDCI, CCI	NDTI > 0.1	Mask turbidity
Emergent vegetation	NDCI, NDVI	MNDWI < 0	Apply water mask
Shallow bottom	NDCI, CCI	Bottom visible in true color	Mask shallow zones
Floating debris	FAI	FDI elevated	Interpret as debris
Sunglint	All	Bright silver patches	Discard scene

False Negatives

Situation	Why Index Misses	What to Do
Subsurface bloom	FAI near zero	Use NDCI + field samples
CDOM-rich water	Visible bands suppressed	Interpret cautiously; use S3 if possible
High turbidity	Sediment overwhelms signal	Wait for settling or use field sensors

Chapter 10

Remote Sensing Caveats

Why color alone is not proof of a bloom and how to avoid over-claims.

Most important caveat

Sentinel-2 does not measure phycocyanin directly. All S2 cyanobacteria mapping is proxy-based. Use S3 for direct PC logic where possible and always field-verify.

Mixed pixels

20 m bands mix land and water. Buffer shorelines by 2–3 pixels before analysis.

Growth phase shifts

Phycobiliprotein peaks can weaken during nutrient stress and stationary phases. Optical signals vary over bloom life cycle.

Toxicity is not optical

Microcystin and other toxins are biochemical and require lab verification.

Chapter 11

Mini Glossary

Concise definitions for key terms.

CDOM

Colored dissolved organic matter that absorbs blue light and makes water appear tea-colored.

Phycocyanin (PC)

Cyanobacterial pigment absorbing ~615–621 nm. Direct satellite retrieval requires Sentinel-3 OLCI.

Red edge

Rapid reflectance rise between red and NIR; used to infer chlorophyll.

Mixed pixel

A pixel containing both land and water, which corrupts index values.

Sunglint

Specular reflection of sunlight that invalidates optical indices.

Field Guide to CopernicusRemote Sensing and Light Indices

Purpose

What Changed

Bottom Line

Operational stack (rivers)

Rapid Cheat Sheet

Sentinel-2 Band Cheat Sheet

Quick Index Table

Aquatic Confounders

Core Concepts

Light as Environmental Data

Photon data

Reflectance

Absorption

Scattering

Red edge

CDOM effect

CDOM and Blue-Loss

What CDOM Does Optically

Absence of blue

Tea-colored optics

Why dark water is not automatically a bloom

How CDOM confounds bloom screening

Copernicus Basics

Sentinel-2 MSI — Key Bands for Water and HABs

Sentinel-3 OLCI — Aquatic Bands

Light Indices Guide

Water / Masking / Aquatic Context

Turbidity / Sediment Context

Bloom / HAB / Cyanobacteria

Vegetation / Red-Edge Context (Confusion Removal)

Field Interpretation

If You See This, Think This

Copernicus Browser Workflow

Step-by-Step

Which Index Should I Use?

Troubleshooting False Positives

False Negatives

Remote Sensing Caveats

Mixed pixels

Growth phase shifts

Toxicity is not optical

Mini Glossary

Field Guide to Copernicus
Remote Sensing and Light Indices