AI Insights

Detecting Automatic Lake Encroachment using Machine Learning and Remote Sensing in Chennai

February 21, 2023


article featured image

Introduction

Encroachment of Water Bodies is due to Urbanisation thus leading to the depletion of natural resources, causing scarcity and people suffering from a lack of food security. The goal of the project is to solve the encroachment issues thereby addressing Urbanization Issues by using Machine Learning and Remote Sensing.

Data Collection

Data is collected from Landsat 7(2000-2014) and Landsat 8(2015-2022) Surface Reflectance products(Level 2) available at Google Earth Engine for the years (2000-2022) with 30 m spatial resolution. 

Preprocessing of Data involves Scan Line Correction which can cause the loss of 22% of data by using Focal mean and blending of Data. The below figure shows the comparison of Scan Line Error present and Removed from Landsat 7 Data.

Fig 1. Composite shows with and without Scan Line Error

Fig 1. Composite shows with and without Scan Line Error

Clouds are Removed in the form of CF Scan Algorithm and Compositing in the form of Yearly Median from Landsat 7 and Landsat 8

Water Body Collection and Classification                         

For the Automation of Water Body Extraction Usage of Machine Learning algorithm such as Random Forest is implemented and classification shows inaccuracies over the data and Expert System based Indices Approach which can be better suited for the extraction of Water Bodies.

Water Based Indices such as Modified Normalized Difference Water Index (MNDWI) is used which is the normalization of  Green(G) and Shortwave Infrared(SWIR) bands for the year of 2000 -2022. MNDWI = (G-SWIR)/(G+SWIR)

Thresholding-based Classification is implemented to extract Water Bodies such as MNDWI > 0  and MNDWI<0  for waterbody and nonwaterbody. Then, Trends in Water Body are Analysed.

Fig 2. Overall Methodology

Fig 2. Overall Methodology

Conclusion

Fig 3. Overall Areas of Lakes from 2000-2022

Fig 3. Overall Areas of Lakes from 2000-2022

Chennai Lakes started to disappear starting from 2000 to 2018 and thus, showing the validation of Encroachment over the years is given in the form of Statistical Analysis in the form of Tables and Graphs.

Fig 4. Encroachment of Pulicat Lake

Fig 4. Encroachment of Pulicat Lake

Fig 5. Encroachment of Puzhal Lake

Fig 5. Encroachment of Puzhal Lake

Fig 6. Encroachment of Retteri Lake

Fig 6. Encroachment of Retteri Lake

Fig 7. Encroachment of Korattur Lake

Fig 7. Encroachment of Korattur Lake

Fig 8. Encroachment of Ambattur Lake

Fig 8. Encroachment of Ambattur Lake

Fig 9. Encroachment of Chembarambakkam Lake

Fig 9. Encroachment of Chembarambakkam Lake

Fig 10. Encroachment of Muttukadu(Great Salt) Lake

Fig 10. Encroachment of Muttukadu(Great Salt) Lake

Fig 11. Encroachment of Madurantakam Lake

Fig 11. Encroachment of Madurantakam Lake

Fig 12. Encroachment of Poondi Lake

Fig 12. Encroachment of Poondi Lake

For Pulicat, Retteri, Madurantakam Northern part of Lakes are encroaching at a  faster rate and for Puzhal, Retteri, Ambattur, Chembarambakkam, Poondi Southern part of Lakes are encroaching also at a faster rate.

This article is written by authors: Sairam K, Biswajit Mondal.

Ready to test your skills?

If you’re interested in collaborating, apply to join an Omdena project at: https://www.omdena.com/projects

media card
Revolutionizing Short-term Traffic Congestion Prediction with Machine Learning
media card
Harnessing AI to Monitor and Optimize Reforestation Efforts in Madagascar
media card
How We Leveraged Advanced Data Science and AI to Make Farms Greener
media card
Interactive Geospatial Mapping for Crime Prevention