Flooding: an urban threat

You have just bought a house in one of the major cities in Australia at a cutting edge price and the real estate agent assured you that it was a bargain deal you could not pass. Now you are standing inside your flooded property wondering how it could have happened. As it turns out, your house is located in a flood zone. Old flood management plans are often outdated, as changes such as increased urbanisation have not been taken into account. Areas that were previously considered to be low risk flooding areas can now be at a much higher risk. So what can government agencies do to maintain more up to date flood management plans?

Decades ago, the use of remote sensing was not as prolific and would still rely on ground measurements, which can be costly and time-consuming. Nowadays, we have access to numerous different remote sensing sensors that can be used on various applications. The access to the vast number of multispectral imaging systems means that the production of high quality maps and the use of land use data makes flood modelling more accessible and cost-effective. Research has shown that a combination of Landsat 7 Enhanced Thematic Mapper Plus (ETM+) and Light Detection and Ranging (LiDAR) sensors can be effective in collecting data that can be used in hydrology models. Both sensors work in different ways. The ETM+ is a radiometer with a passive sensor, which means that it measures reflection of electromagnetic radiation off surfaces. LiDAR, on the other hand, is an active sensor, meaning that it emits laser pulses and measures the time it takes for these laser pulses to return to the sensor – the longer it takes for a pulse to return, the less the elevation of the object that reflected the pulse is on Earth. The difference between the two sensors makes them particularly useful for different purposes.

0027_DAI6 (2)
Passive vs active sensors (source)

The way various substances and objects reflect and absorb electromagnetic radiation is referred to as a spectral signature. Water, for example, absorbs near-infrared radiation, unlike soil and green vegetation. Using this unique feature of water’s spectral signature means that data from a sensor such as the EMT+ is able to be classified (sorted into different groups). For the purpose of creating a flood model, being able to classify features into vegetation/soil, urban areas and water bodies is of key importance as these features are used differently in a model. The former two are important to know to determine how well water would be absorbed by the surface, also called the imperviousness of a surface. Developed, urban areas are more prone to higher surface runoff due to the increased imperviousness of the surface. Concrete simple cannot absorb as much water was a forest and this is a big issue with increased urbanisation across the globe. The collected data can be compared to previous years to create a time-series, which allows for the identification of changes. To actually use this information to determine in what direction water could potentially flow, the topography of the surface should be known.


The different spectral signatures of water, soil and green vegetation (source)

While EMT+ is good to use to identify different feature classes, LiDAR’s usage of laser pulses and high vertical accuracy makes it an effective tool to measure topography – or elevation differences – of a surface. Last return pulses are generally an indication of the surface, as it would be the furthest away from a sensor. Using a vast number of return values and filtering out only the points that are ground values, a digital elevation map (DEM) can be created. As water generally flows from higher to lower elevations, the direction runoff will flow towards and where it could pool up can be determined.

Combined, the two sensors provide a vast amount of data that is comprehensive and can be used in numerous of hydrology models. Adding rainfall data, obtained through radar and classic tools such as rain gauges, even simple flood models will be able to show under which circumstances flooding may occur and what areas would be mostly affected. Additionally, the processing and modelling of data can be done in a very short amount of time, making it possible to provide frequently updated flood maps. The vast database of LiDAR DEM data and ETM+ data provided for the public to use means that you do not necessarily have to fly a small plane or unmanned aerial vehicle (UAV) equipped with LiDAR, bringing the costs of data acquisition down. The accessibility of remote sensing data allows various types of people to tinker with it, using older data as calibration for new models or developing other applications, such as early-warning systems to safeguard people’s lives or frequently updated maps of flood prone areas.

Remote sensing is an effective and quick method of creating flood models. The accuracy and accessibility of both ETM+ and LiDAR make the combination of both remote sensing techniques an accessible datasource for flood prediction and floodplain management by various government agencies. With the advancement of mobile technology and further research into remote sensing algorithms, the ability pull up real-time flood data on your phone and call your real estate agent out on a lie (or what most people would do – check whether you are safe after heavy rain) could be a real possibility in the near future!

*Featured image depicting Zeeland, The Netherlands, obtained from the European Space Agency website

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s