FloodGauge is a portal that aims to estimate the probability of flooding in a region based on its history of changing land cover and an increase in demand for water and forest resources owing to the increasing population. By making timely policy changes and ensuring organized infrastructural development, cities can reduce the chances of flooding. As part of our proposed future work, the model will be able to predict floods at least a week in advance to help citizens and authorities execute the timely evacuation process. Cities that understand the urgency of action, upgrade their drainage systems, and ensure stricter laws against land reclamation stand the highest chance of preventing flooding or at least reducing the losses.

How does it work?

Different cities have different complexities and connectivity of drainage systems. It is hard to quantify the extent of connectivity and effectiveness of these systems. However, land reclamation is one factor that is increasingly making cities prone to flooding. This is particularly what we study. History of EO land reclamation, forest cover, and water level data in flood-affected regions can help forecast the likelihood of areas being prone to flood based on their increasing land cover patterns. 

The steps to draw correlations with land cover and forecast floods include the following:

1. Sentinel-1 SAR images are used to get the land cover backscatter classifications with the help of Sentinel-2 optical images using a supervised classifier Pixel-based change detection is used for detecting variations in water levels prior to and post-flooding. Backscatter characteristics and variation rules of different ground objects are essential prior knowledge for flood analysis.
2. Flooding intensity is modeled on the water level, land cover, and precipitation data
3. We obtain the history of flood intensity maps over several years for an input geolocation using step1 and compare the effect with the variation of deforestation and human settlements over years .

Additionally, a growing population is increasing the demand for forest and water resources. As a future step, we will correlate population growth with land reclamation and deforestation. This will help policymakers drift to alternative sources to fulfill demands.

Insights

We correlate the deforestation and land cover patterns along with the history of flooding in 5 Indian cities and infer the following:

1. Bihar: Although the state is developing its infrastructure, large fractions of the state continue to be farming lands, more than most unaffected states in India. However, Bihar has been consistently affected by flooding from the Kosi river on the east. Bihar is one of the states where the land reclamation is not a significant factor, yet infrastructural reforms are required to mitigate losses. One key idea in this respect is to help citizens migrate from the coast to permanent settlements in safer regions of the city with well-built drainage systems.
2. Mumbai: Over the last century, Mumbai has reclaimed large fractions of the Arabian Sea to develop infrastructure, more so in the last two decades. The industrial hub is also home to an ever-increasing massive population. The existing drainage systems are inefficient and unable to channel rainwater into the ground/sea. This results in extensive flooding in the city resulting in shutdown of offices, schools, and consequent loss of lives and economy.
3. Chennai: Land reclamation and building infrastructure over lakes has reduced natural sinks and groundwater in the last two decades. This has necessitated a better drainage system which has not been built to cope with the rainfall. Accordingly, Chennai experiences floods every year leading to severe loss of lives and property. Additionally, the reduction in groundwater has led to significant shortage of freshwater in the city in the last decade with thousands succumbing to the heat without adequate access to drinking water. The case of the Pallikaranai marshlands, which drains water from a 250-square-kilometre catchment, is telling. Not long ago, it was a 50-square-kilometre water sprawl in the southern suburbs of Chennai. Now, it is 4.3 square kilometres – less than a tenth of its original. The growing finger of a garbage dump sticks out like a cancerous tumour in the northern part of the marshland. Two major roads cut through the waterbody with few pitifully small culverts that are not up to the job of transferring the rain water flows from such a large catchment. The edges have been eaten into by institutes like the National Institute of Ocean Technology. Ironically, NIOT is an accredited consultant to prepare Environmental Impact Assessments on various subjects, including on the implications of constructing water bodies. The vast network of water bodies that characterised Chennai can only be seen on revenue maps now. Of the 16 tanks belonging to the Vyasarpadi chain downstream of Retteri, none remain.
4. Delhi: Depending upon the river flow level down stream, it takes about 48 hours for Yamuna level in Delhi to rise. The rise in water level causes a backflow effect on the city's drains. The city also experiences floods due to its network of 18 major drains having catchment areas extending beyond the city's limits. In Delhi, the Yamuna's floodplain is defined as that area near the river that gets submerged at least once in 25 years. Rampant unchecked construction on the floodplains of the Yamuna poses a grave threat. Almost 30 percent of the floodplain in the city has already been compromised and is no longer available to the river. The industrial hubs of Gurgaon and Noida attract millions of youth for job prospects. The industrial activities and pollution in the city have impaired Yamuna which was otherwise a natural sink. While the poor drainage planning is something to improve, the population spread is also to be accounted for in Delhi. 
5. Assam: Assam is a flood-prone state with its vast network of rivers. A combination of natural and man made factors have contributed to the flood-related devastation in the state. In the monsoon period each year, Brahmaputra and Barak River with more than 50 numbers of tributaries feeding them, cause the flood devastation Another major problem is the bank erosion caused by the river Brahmaputra, Barak, and its tributaries in the state. This causes water to overflow in the low-lying areas every time a flood occurs.Last but not the least, a major man made problem that has affected the state of Assam is the building of embankments which started from the 1950s. Unplanned construction of these embankments have disrupted Brahmaputra’s natural flow, studies have revealed.

Impact

Securing lives and property during floods is a temporary solution. In many cases, particularly in developing countries, even this basic management fails. Prevention of floods is the need of the hour. Reckless infrastructural activities by humans and poor policy reforms, coupled with poor drainage management by developing authorities are leading issues to address. Our model helps quantify the need for these reforms and qualitatively suggests region-specific steps for flood prevention. From a wider perspective, population growth can be viewed as one root cause of increased settlements and deforestation. Governments can use our results to understand the demographics and demand in their countries, and thus initiate policies to promote the optimal spread of the population (by developing states uniformly thus preventing biased attraction for migration to particular regions)