Land Use Land Cover (LULC) — Bhuvan/NRSC

India's national land mapping — track how land use has changed from 2005 to 2023 at 1:50,000 scale.

---

Provider

NRSC / ISRO (via Bhuvan)

Website

bhuvan.nrsc.gov.in

Access Level

🟡 Free Registration on Bhuvan

Formats Available

GeoTIFF (raster), Shapefile (vector), WMS

Coverage

All India — 1:50,000 scale (~56m resolution

Time Points

2005, 2011, 2015, 2017, 2019, 2021, 2023

---

What Is LULC?

Land Use Land Cover (LULC) maps show what the land surface looks like — whether it's a forest, a paddy field, a city, a water body, or a wasteland. NRSC produces LULC maps for all of India using satellite images from ResourceSat (LISS-III and LISS-IV sensors).

Having LULC at multiple time points (2005–2023) allows you to detect: - Urban expansion eating into agricultural land - Forest degradation or recovery - Wetland conversion - Agricultural intensification (double/triple cropping)

---

LULC Class System

NRSC uses a hierarchical legend with 3 levels of detail:

Level 1 (Broad)	Level 2 (Detailed)	Level 3 (Specific)
Agriculture	Kharif cropland	Paddy, Jowar, Bajra...
	Rabi cropland	Wheat, Mustard...
	Double/Triple crop	Irrigated, Unirrigated
Forest	Dense Forest	>70% canopy
	Open Forest	10-40% canopy
	Mangroves	Coastal
Wasteland	Degraded	Various types
	Barren	Rocky, sandy
Built-up	Urban	Cities, towns
	Industrial	Factories, mines
Water Bodies	Reservoirs	Dams
	Rivers	Perennial, seasonal
	Ponds	Village tanks
Wetlands	Inland	Lakes, swamps
	Coastal	Estuaries, tidal flats

The full 48-class legend is available in the NRSC LULC technical document on bhuvan.nrsc.gov.in.

---

How to Download LULC Data

Step 1: Register on Bhuvan

1. Go to bhuvan.nrsc.gov.in → Register
2. Choose "Research" as purpose of use

Step 2: Download State/District LULC

1. After login: Go to "Thematic Data" → "Land Use"
2. Select Year (2005, 2011, 2015, 2017, 2021, 2023)
3. Select State
4. Download GeoTIFF (raster) or Shapefile (vector)

Step 3: Open in QGIS

1. Drag the downloaded GeoTIFF into QGIS
2. Right-click layer → Properties → Symbology
3. Change to Paletted/Unique Values → Classify
4. Each pixel value (1–48) represents a land class
5. Download the NRSC colour scheme file (.qml) from Bhuvan for correct colours

---

Python: Calculate Land Class Areas

import rasterio
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt

# LULC legend (simplified)
LULC_CLASSES = {
    1: 'Kharif Cropland',
    2: 'Rabi Cropland',
    3: 'Double/Triple Cropland',
    4: 'Current Fallow',
    5: 'Other Fallow',
    6: 'Dense Forest',
    7: 'Open Forest',
    8: 'Scrub Forest',
    9: 'Mangroves',
    10: 'Built-up Urban',
    11: 'Built-up Industrial',
    12: 'Cropland + Trees',
    13: 'Plantations',
    14: 'Grassland/Grazing',
    15: 'Wetland',
    16: 'Water Body',
    17: 'Barren/Unculturable',
    18: 'Shifting Cultivation',
}

def calculate_lulc_areas(tif_path: str) -> pd.DataFrame:
    """Calculate area of each LULC class in km²."""
    with rasterio.open(tif_path) as src:
        data = src.read(1)
        # Pixel size in degrees (approximately 0.0005° ≈ 56m at equator)
        pixel_area_km2 = abs(src.transform.a * src.transform.e) * 111 * 111

    unique, counts = np.unique(data[data != src.nodata if hasattr(src, 'nodata') else data != -9999],
                                return_counts=True)
    df = pd.DataFrame({'class_id': unique, 'pixel_count': counts})
    df['area_km2'] = df['pixel_count'] * pixel_area_km2
    df['class_name'] = df['class_id'].map(LULC_CLASSES).fillna('Other')
    df = df.sort_values('area_km2', ascending=False)
    return df

# Analyse 2005 and 2023 LULC
lulc_2005 = calculate_lulc_areas("maharashtra_lulc_2005.tif")
lulc_2023 = calculate_lulc_areas("maharashtra_lulc_2023.tif")

# Compare
merged = lulc_2005.merge(lulc_2023, on='class_name', suffixes=('_2005', '_2023'))
merged['change_km2'] = merged['area_km2_2023'] - merged['area_km2_2005']
merged['change_pct'] = (merged['change_km2'] / merged['area_km2_2005']) * 100

print("\nLand Use Change 2005 to 2023 — Maharashtra")
print(merged[['class_name', 'area_km2_2005', 'area_km2_2023', 'change_pct']].to_string())

---

✏️ Practice Exercise

Exercise 3.2 — Urban Expansion Analysis

Goal: Calculate how much urban area has grown in your district from 2005 to 2023.

1. Download LULC 2005 and LULC 2023 GeoTIFF for your state from Bhuvan
2. Open QGIS → Load both files
3. Clip both to your district using Raster → Extraction → Clip Raster by Mask Layer
4. Use Raster → Raster Calculator:
5. Expression: ("lulc_2023@1" = 10) - ("lulc_2005@1" = 10)
6. This gives +1 where urban grew, -1 where it shrank, 0 where unchanged
7. Count pixels using Raster → Analysis → Raster Layer Statistics

Or in Python:

import rasterio
import numpy as np

with rasterio.open("district_lulc_2005.tif") as s1:
    lulc05 = s1.read(1)
    pixel_area = abs(s1.transform.a * s1.transform.e) * 111*111  # km²

with rasterio.open("district_lulc_2023.tif") as s2:
    lulc23 = s2.read(1)

urban_2005 = np.sum(lulc05 == 10) * pixel_area
urban_2023 = np.sum(lulc23 == 10) * pixel_area

print(f"Urban area 2005: {urban_2005:.1f} km²")
print(f"Urban area 2023: {urban_2023:.1f} km²")
print(f"Increase: {urban_2023 - urban_2005:.1f} km² ({((urban_2023-urban_2005)/urban_2005)*100:.1f}%)")

---

Next Dataset: Forest Cover — FSI →