🌾 Remote Sensing and Crop Monitoring

Remote sensing gives scalable crop intelligence by converting reflectance patterns into actionable signals for growth, stress, and yield prediction.

---

Principles of Remote Sensing

Remote sensing is the science of acquiring information about objects or areas from a distance, typically using satellites or aircraft, without physical contact. It works on the principle of electromagnetic radiation (EMR) interaction with Earth's surface.

Every object on Earth:

• Reflects some portion of incoming radiation
• Absorbs some portion
• Transmits some portion (primarily for liquids)

The ratio of reflected energy to incoming energy is called reflectance. Remote sensing sensors measure this reflectance across different wavelength bands.

---

Electromagnetic Spectrum in Remote Sensing

Band	Wavelength Range	Agricultural Use
Visible – Blue	400–500 nm	Chlorophyll absorption, water body delineation
Visible – Green	500–600 nm	Vegetation green peak reflectance
Visible – Red	600–700 nm	Chlorophyll absorption (key for NDVI)
Near Infrared (NIR)	700–1300 nm	Strong leaf cell structure reflectance
SWIR (Short Wave IR)	1300–2500 nm	Soil moisture, canopy water content
Thermal IR	8,000–14,000 nm	Land surface temperature, crop water stress
Microwave (SAR)	1 mm–1 m	All-weather, cloud penetration; soil moisture

The NIR region is the most important for vegetation studies — healthy leaves have very high NIR reflectance (up to 50%) due to the spongy mesophyll cell structure.

---

Spectral Signatures

A spectral signature is the characteristic pattern of reflectance across wavelengths for a specific surface type. It acts as a "fingerprint."

Surface	Red Reflectance	NIR Reflectance	Behaviour
Healthy crop	Low (chlorophyll absorbs)	High (mesophyll reflects)	Low Red, High NIR
Stressed crop	Higher (less chlorophyll)	Lower (wilting reduces NIR)	Higher Red, Lower NIR
Bare soil	Moderate	Moderate	Gradual increase with wavelength
Water body	Very low	Near zero (absorbs NIR)	Dark in NIR

These differences in spectral signatures form the basis for crop type mapping and stress detection.

---

Vegetation Indices

Vegetation indices are mathematical combinations of spectral bands that highlight vegetation properties and reduce effects of soil background, atmosphere, and sun angle.

NDVI (Normalized Difference Vegetation Index)

$NDVI = \frac{NIR - Red}{NIR + Red}$

• Range: -1 to +1
• Values: <0 = water/cloud; 0–0.2 = bare soil/rock; 0.2–0.4 = sparse vegetation; 0.4–0.6 = moderate crops; >0.6 = dense, healthy vegetation
• Most widely used index for crop growth monitoring, yield forecasting

EVI (Enhanced Vegetation Index)

$EVI = 2.5 \times \frac{NIR - Red}{NIR + 6 \times Red - 7.5 \times Blue + 1}$

• Reduces atmospheric effects and soil background noise
• Better performance in high-biomass dense canopy areas where NDVI saturates

SAVI (Soil-Adjusted Vegetation Index)

$SAVI = \frac{(NIR - Red)(1 + L)}{NIR + Red + L}$

Where L = soil correction factor (typically 0.5). Better than NDVI in sparse vegetation or semi-arid regions where soil background is prominent.

NDWI (Normalized Difference Water Index)

$NDWI = \frac{Green - NIR}{Green + NIR}$

• Monitors water content in vegetation canopy and surface water bodies
• Useful for drought monitoring and irrigation management

NDRE (Normalized Difference Red-Edge Index)

$NDRE = \frac{NIR - RedEdge}{NIR + RedEdge}$

• Uses the red-edge band (700–730 nm), highly sensitive to chlorophyll content
• Excellent detector of early nitrogen stress before visible symptoms appear
• Available on Sentinel-2 (Band 5) and most commercial multi-spectral UAV sensors

---

Key Satellite Platforms

Indian Satellites

Satellite	Sensor	Resolution	Swath	Use
Resourcesat-2/2A	LISS-III	23.5 m	141 km	Crop mapping, LULC
Resourcesat-2/2A	LISS-IV	5.8 m	70 km	Field-level monitoring
Resourcesat-2/2A	AWiFS	56 m	740 km	State-level crop inventory
Cartosat-3	PAN	0.25 m	16 km	High-res land records
RISAT-1A	SAR (C-band)	3–50 m	Variable	Cloudy-season crop mapping

RISAT-1A (Radar Imaging Satellite) uses Synthetic Aperture Radar (SAR) — it can image through clouds and rain, critical during the Kharif season when cloud cover prevents optical imaging.

Global Satellites

Satellite	Agency	Resolution	Bands	Data Access
Landsat 8/9	NASA/USGS	30 m (OLI)	11 bands	Free (USGS Earth Explorer)
Sentinel-2A/2B	ESA	10 m (RGB, NIR), 20 m (SWIR, Red-edge)	13 bands	Free (Copernicus Hub)
MODIS	NASA	250 m–1 km	36 bands	Free; daily coverage
VIIRS	NOAA/NASA	375 m	Multiple	Free; fire/crop stress

Sentinel-2 with its 10 m resolution and 5-day revisit time has become the most widely used satellite for precision agriculture globally.

---

Agricultural Remote Sensing Applications

1. Crop Area Estimation

• Multi-temporal NDVI time series → crop phenology profiles
• Machine learning classification of satellite imagery → crop type maps
• Example: ISRO's AWiFS-based national wheat, rice, and soybean acreage estimation

2. Yield Forecasting

• NDVI at critical growth stages (tillering, flowering) correlates with final yield
• Integrated with agro-meteorological models (temperature, rainfall) → district-level yield forecast

3. Drought Monitoring

• NDWI and Vegetation Condition Index (VCI) track moisture stress
• Combined with rainfall anomaly data (IMD, CHIRPS) for drought severity mapping

4. Crop Disease and Pest Detection

• Stressed crops show elevated red reflectance + reduced NIR
• Hyperspectral sensors (400–2500 nm, 100+ bands) detect subtle biochemical changes
• Early detection before visual symptoms → timely intervention

---

FASAL Project

FASAL (Forecasting Agricultural output using Space, Agro-meteorology and Land based observations) is India's operational remote sensing-based crop forecasting system operated by ICAR, IMD, SAC (ISRO), and DES (MoA&FW).

• Pre-harvest forecasts for 10 major crops (wheat, rice, soybean, cotton, etc.)
• Uses: AWiFS satellite data + agro-meteorological models + ground observations
• Provides district-level crop area and yield estimates before harvest
• Aids government procurement policy, price stabilization, and food security planning

Google Earth Engine (GEE)

Google Earth Engine is a cloud-based geospatial analysis platform providing:

• Access to 40+ years of satellite imagery archive (Landsat, Sentinel, MODIS, etc.)
• Parallel computing for planetary-scale analysis
• Free for research and education
• JavaScript and Python APIs

Use case: Compute 20-year NDVI trend for a district to detect agricultural land degradation.

---

Remote sensing transforms millions of hectares of farmland into digitally monitored landscapes — enabling data-driven policy, early warning systems, and precision input management at scales impossible with ground surveys alone.

---

Summary Cheat Sheet

Topic	Key Point
Core principle	EMR interaction and spectral reflectance mapping
High-value bands	Red, NIR, SWIR, thermal, and microwave
Key indices	NDVI, EVI, SAVI, NDWI, NDRE
Operational use	Area estimation, stress detection, and yield forecasting

References

3 sources

ISRO and NRSC crop monitoring and remote sensing resources.

Sentinel and Landsat mission documentation for agri-use cases.

ICAR publications on vegetation indices and crop forecasting.