The Power of Sentinal-2 Data and Google Earth Engine in Flood Area Mapping(Haridwar district)
Haridwar (Uttarakhand), July 12: Continuous rainfall and a breach in the Sonali River dam have raised a threat of floods in the Laksar town of Uttarakhand’s Haridwar district. The Sonali River dam broke near the village of Kuan Kheda, according to officials.
The joint teams of State Disaster Response Force (SDRF), National Disaster Response Force (NDRF) and Indian Army have rescued over 550 people from areas facing flood-like situations in Haridwar district.
According to the State Emergency Operation Centre (SEOC), more than 50 areas in the district are heavily waterlogged, with villages in Laksar and Roorkee tehsil among the worst-hit.
I am using Sentinal image for Flood area mapping that particular time using Google Earth Engine
The multispectral nature of Sentinel-2 images (13 spectral bands) allows for the discrimination between water and land surfaces, vegetation, and urban infrastructure. Specifically, the near-infrared (NIR) and short-wave infrared (SWIR) bands are instrumental in detecting water, as water bodies absorb most of the sunlight in these wavelengths, making them appear distinctly in the imagery. This property is particularly useful for identifying and delineating floodwaters, even in areas with dense vegetation.
Methodology:
Step 1: Import Required Dataset
var collection = ee.ImageCollection("COPERNICUS/S1_GRD")
.filterBounds(table)
.filter(ee.Filter.listContains("transmitterReceiverPolarisation","VV"))
.select('VV')Step 2: Display the Region of Interest
// Display the Region of Interest
var roi_style = {
color: "red",
fillColor: "00000000",
width: 1.5
};
Map.addLayer(roi.style(roi_style), {}, "Region of Interest");
Map.centerObject(roi, 10);Step 3: Prepare Sentinel-2 Imagery
// On 15 july 2023, severe floods affected theHaridwar District of Uttarakhand
// Select images by predefined dates
var before = collection.filterDate("2023-06-01","2023-06-15").mosaic()
var after = collection.filterDate("2023-07-14","2023-07-15").mosaic()
var before_clip = before.clip(roi)
var after_clip = after.clip(roi)
Step 4: Apply smoothening Filtering :
var before_s = before_clip.focal_median(30,"circle","meters")
var after_s = after_clip.focal_median(30,"circle","meters")Step 5: Calculate Differencing :
// difference
var difference = after_s.subtract(before_s)
Step 6: Apply a Threshold:
var flood_extent = difference.lt(-3)
var flood = flood_extent.updateMask(flood_extent)
// Display the Maps
Map.addLayer(before_clip,{min:30,max:0},"before_flood")
Map.addLayer(after_clip,{min:30,max:0},"after_flood")
Map.addLayer(difference,{},"difference")
Map.addLayer(flood,{},"flood")
Conclusion:
The integration of Sentinal-2 data, the Refined Lee algorithm, and GEE offers a comprehensive approach to flood management. It equips us with the means to mitigate the impact of flooding and respond to emergencies.
In conclusion, the power ofSentinal -2 data and GEE in flood area mapping transcends boundaries and overcomes obstacles. It is a testament to human ingenuity and technological advancement. As our world faces the growing challenges of climate change and the increasing frequency of extreme weather events, these tools become ever more critical.
