Disaster management needs constellations of satellites with multispectral sensors, says Indian space researcher, Ranganath Navalgund.
Ever since 1972, when the United States launched the first Earth observation satellite, Landsat-1, data from satellite remote sensing has identified disaster sites, helped assess damage and risks, monitored situations and offered early warnings.
But disasters come in all shapes and sizes, needing varying data during the disaster cycle of mitigation, preparedness, response and recovery. No single satellite can hope to meet all these needs. Rather, what disaster managers need is a constellation of satellites carrying a range of sensors.
Crucially, different situations need data collected in different wavebands.For example, optical and near infrared data can map land use or assess agricultural droughts. But to track a cyclone's eye, or monitor flooded areas beneath cloud, microwave sensors are needed.
And landslide studies depend on accurate high-resolution digital elevation models, which require data collected by stereo-viewing optical sensors (e.g. Cartosat – 1), Interferometric Synthetic Aperture Radars (InSARs) or Light Detection and Ranging (LIDAR) instruments.
Then again, for fires or volcanoes, it is thermal imagery that is needed — to pick up hotspots.
Disaster managers really need satellites incorporating sensors that collect data in all these regions of the electromagnetic spectrum.
An awkward trade-off
There is often an awkward trade-off between temporal and spatial resolutions. Managing many natural disasters, such as cyclones or fires, demands detailed and continuous data.
But although geostationary satellites, such as Meteosat or INSAT (Indian National Satellite System)/ Kalpana, provide almost constant surveillance (every 15 minutes), they lack detail (their spatial resolution is low). Conversely, polar-orbiting satellites offer higher spatial resolution data (even down to less than 1m) but information is only collected once every few days.
In flood monitoring, this can pose a real problem. Low spatial resolution data can map out large inundated areas, but relief efforts really need more detailed, yet still timely, data on infrastructure, like submerged bridges, drains and roads.
A constellation of polar-orbiting satellites, equally spaced around a sun-synchronous orbit to provide continuous coverage over any given place, could solve this.
Such a constellation, designed primarily for disaster management, could offer more frequent data in the right part of the spectrum.
For example, geostationary satellites, predominantly designed for weather forecasting, are good at spotting a cyclone as it's forming, tracking its movements, and predicting land fall points. But they don't usually carry microwave sensors, which are needed to estimate a cyclone's intensity — critical for predicting potential damage.
Exactly how many satellites an effective constellation would need is still open for debate. But many studies suggest at least eight, with dual capability sensors that can collect both high and low spatial resolution data,
and an equal split between optical (including thermal) and microwave instruments.
The satellites should also be agile, i.e. they should allow rapid changes in camera orientation so a disaster area can be kept in view longer.
One such Disaster Monitoring Constellation (DMC) has been designed by Surrey Satellite Technology Ltd., UK. Seven equally-spaced satellites orbit at 686 km, with 26-32 m resolution and multi-spectral capacity.
Each satellite is independently owned and controlled by a separate nation, but all are committed to being a collaborative network and the satellites have been spaced so that the network as a whole can provide daily images. However, the satellites in this DMC do not have sensors in thermal and microwave region, limiting their capabilities.
A satellite constellation that does offer microwave images is SAR-Lupe of Germany. This military reconnaissance system comprises five satellites that operate in three 500-kilometre orbits roughly sixty degrees apart. They use the X Band segment of microwave radiation, providing very high resolution (< 1 m) data. Such data could be very useful for man-made disasters.
Back to Earth
Of course, the most important thing for effective disaster monitoring and mitigation is that satellite data reaches managers and emergency planners in an easy-to-use format. So satellites in a constellation must provide a certain amount of on-board processing and automatic analysis(as for the MODIS fire map, (see Remote sensing for disasters: Facts and figures).
Proper communication infrastructure that gets data from the satellite to the end-user is also crucial. Combining remote sensing satellites with communication satellites is useful. For example, India combines its Indian Remote Sensing satellite system (IRS) — designed for land use and ecological monitoring — with the INSAT (Indian National Satellite System) communications satellites.
Being able to integrate satellite data with other geo-spatial datasets and environmental models is also crucial. Satellites aren't the whole answer. To assess landslide risk, for example, you must integrate remote sensing data with population maps and other spatial databases.
Similarly, to forecast or warn of floods you need real-time satellite data on rainfall intensity and river discharge, but also in-situ observations, knowledge of the topography, and hydrological models.
Ranganath Navalgund is director of the Space Applications Centre at the Indian Space Research Organisation in Ahmedabad, India.