Monsoon Overflow - Space Apps Challenge

Monsoon Overflow| A Flood of Ideas

Vallabh Vidhya Nagar, Anand

Awards & Nominations

Monsoon Overflow has received the following awards and nominations. Way to go!

Best Use of Data

The solution that best makes space data accessible, or leverages it to a unique application.

Local Peoples' Choice Winner

A Flood of Ideas

Your challenge is to develop a new methodology or algorithm that leverages Earth observation and critical infrastructure datasets to estimate damages to infrastructure caused by flooding. Make a measurable impact on the resilience of nations by helping the Earth observations community contribute to the United Nations’ primary effort to reduce disaster risk!

Critical Infrastructure Damage Detection Using Satellite Images and Crowd Sourced Data

Summary

Floods are the most frequent natural disaster occurs when an overflow of water submerges land due to excessive rain, ruptured dam, rapid melting of snow or Tsunami. Floods can cause widespread devastation, resulting in loss of life, damages to property and critical infrastructure. The infrastructure are prone to damage due to their geographical extent is a determinant of, and is determined by, patterns of human development, which is often concentrated in floodplains. By using Satellite and crowd-source data the detection of this critical infrastructure damage can be carried out for timely decision making.To achieve an Ultimate Goal: Zero casualty and Quick recovery (within 24 hours !!!)

How We Addressed This Challenge

Project Summary:

Our Project has two parts, a mobile application named InfraSave which is used to collect data including geotagged image from the public, about critical infrastructure damage caused due to flood, and in second part we have developed the Damage Detection Algorithm which is used to get the extent of damage in Critical Infrastructure using satellite pre flood and during/post flood images.

Importance of Project:

From 1998 to 2019, floods affected more than 2.5 billion people worldwide. Floods are getting worse because of the climate crisis, decisions to populate high-risk areas and land sink age from the overuse of groundwater. So flood has become a recurring event in many parts of world. Flood causes damage to property and leads to loss in millions of dollar every years.

To timely identify and repair the critical infrastructure is necessary for social and economic functioning of a community or society affected by flood. It is important that authority knew about the extent of damaged flood has caused to the infrastructure and then takes action to restore the critical infrastructure for fast restoration of normal routine and hassle free life of flood affected community.

In this Project, We used crowd source data and satellite images for damage detection, the damage to the critical infrastructure can be identified by the decision makers from the control room and utilize all available resources for efficient working. Work can be prioritized as per the real time situation of the flood in that area. Also decision maker can use the data provide by the Mobile Application if Spatiotemporal Satellite Image of that area is not available or unclear due to cloudy atmosphere.

Methodology of Project:

Whenever User feed the data in mobile application InfraSave i.e. Type of Infrastructure, Observers Name, damage extent, water level, description, Location using GPS, Geotag photo related to damage occurred in infrastructure due to flood, the data are sent to the server and stored in database. The database is linked to the dashboard for authority, where that data can be visualized. Then authority applies the Satellite Image Damage Detection Algorithm on the location reported by crowd source data and find out the extent of damaged caused in that particular structure

In this newly developed algorithm, ESA Sentinal-2 pre flood and post flood images of spatial resolution of 10 meter is used. Due to medium resolution images structures are not clearly visible. Still by Visualization of True Color Composite, False Color Composite, Short Wave Infrared Band few features can be identified. So algorithm we developed uses the Pan Sharpening filter/ High Pass Filter for edge detection of Infrastructure. The filter images give the sharp edge of all the building footprints, roads, railways and other critical Infrastructure. The pre and post flood image can be compared using the slider widget to get the extent of damage in that structure.

Case Study: Bihar State, India- Flood in June 2020

Bihar is India's most flood-prone state, with 76% population in the North Bihar living under the recurring threat of flood devastation. Bihar makes up 16.5% of India's flood affected area and 22.1% of India's flood affected population. About 73.06% of Bihar's geographical area, i.e. 68,800 square kilometers out of 94,160 square kilometers is flood affected. On an annual basis, floods destroy thousands of human lives apart from livestock and assets worth millions.

In June 2020, due to excess rainfall several districts i.e. Sitamarhi, Sheohar, Supaul, Kishanganj, Darbhanga, Muzaffarpur, Gopalganj, West Champaran, East Champaran, Khagaria, Saran and Samastipur were submerged in water for severals days. The flood has damaged various buildings, roads, bridges, railway lines, electrical poles, health care centers, education buildings, etc.

Due to large flooded area and unavailability of location data related to critical infrastructure especially in village areas it is difficult task to identify the critical infrastructure damage without visiting the sites, which leads to delay in work of recovery to flood affected community. So, Critical Infrastructure Damage Detection Using Satellite Images algorithm was used to estimate the damaged caused by flood to critical infrastructure.

Benefits of the Project:

Better Decisions Making = Satellite Image Damage Detection Algorithm + Crowd Source Data

• Crowd source data can be used for fast response due to unavailability of spatiotemporal satellite data of flood prone area.

• Real time validation of data using crowd sourcing.

• Statistical data for quick decision making.

• Authority have a scientific and statistically data for decision making.

• Timely action will be taken to save critical Infrastructure.

• In case of limited resource/re-enforcement priority can be given to most important infrastructure.

How We Developed This Project

Inspiration for Selecting this Project:

Almost Every year many countries in world faces the urban flood as pervious layers in area are converted into impervious layer which leads to high rainfall runoff and results in flooding or due to excess rainfall or rapid slow melting results in overflow of river water leads to flooding.

Due to flooding, many people lost their lives, damage to property, destruction of crops, loss of livestock, and deterioration of health conditions owing to waterborne diseases. As communication links and infrastructure such as power plants, roads and bridges are damaged and disrupted, economic activities come to a standstill, people are forced to leave their homes and normal life is disrupted.

Damage to infrastructure also causes long-term impacts, such as disruptions to supplies of clean water, wastewater treatment, electricity, transport, communication, education and health care. Loss of livelihoods, reduction in purchasing power and loss of land value in the floodplains can leave communities economically vulnerable.

Few countable countries have advanced space technology which can be used for better disaster management while undeveloped or developing countries lacks in space technology related infrastructure /knowledge/ facility So by using NASA, ESA, etc. space agency open source data, undeveloped /developing country can also give better response in disaster management.

Approach to Achieve the Outcome:

We considered holistic approach- Bottom to Top approach for developing this project.

Due to flood, affected people had to suffer for few days to few weeks as basic necessity infrastructure is disrupted such as drinkable water supply, electricity, school, public health centers, etc. Due to unavailability of ground base data timely action are not taken and sometime due to limited resources it difficult to save critical infrastructure which are very important for social and economic growth of society.

Authority does not have real time data validation system so the decision making is delayed and leads to more trouble to public affected by flood. So by using Geospatial Technology, Satellite Images, GPS Location along with Crowd Sourcing data collection method leads to save time, money and critical infrastructure for sustainable growth on earth for mankind.

As now a day Smart Mobile is common and used everywhere by almost everyone, In absence of Satellite Image, Crowd Sourced Data will be very useful. Using Android Studio, Java and Kotlin language InfraSave Mobile Application was developed for data collection from users related to damaged infrastructure; the data is stored in MySQL database and can be visualized by authority.

Using Google Earth Engine, the decision making dashboard/Application was created to run the Damage Detection Algorithm on the pre and post flood Sentinal-2 Satellite Images on affected area using the location provide by the Mobile Application InfraSave user, stored in database. Here JavaScript was used to develop the application in GEE.

Tools / Coding Languages/ Software Used:

•Java

•Java Script

•Kotlin

•PHP

•Android Studio

•MySQL

•Google Earth Engine

Problem Faced:

• Availability of High Spatial (less than 1 meter) and Temporal Resolution Satellite Images during the flood time for more accurate results.

• Optical Remote Sensing data has high cloud coverage so atmospheric correction is required on datasets.

• The data collected by crowdsourcing should be accurate and should be able to well define the problem so that critical infrastructure damage can be identified and repaired as soon as possible.

Achievements:

• By using this technique real time action can be taken by authority without physically visiting the site and can also find the extent of damage caused in area using satellite images.

• The crowd source data will be useful to get details and close range photo about the critical infrastructure damage in the flooded area.

• Satellite Damage Detection Algorithm will be comparing the pre and post flood image and will give the damage occurred in the critical infrastructure due to flooding.

How We Used Space Agency Data in This Project

Data Used:

ESA Sentinal-2 Images: Pre and Post flood Sentinel Images are the Input given to the algorithm to get the damage occurred in critical infrastructure due to floods.