Earth observation (EO)

Remote sensing in combination with advanced analysis techniques have become a decision-supporting tool for humanitarian professionals due to its observational power, ubiquitous usage and availability. Remote sensing and related derived information products are complementary data sources to field-based surveys enriching the pool of spatially aware technologies for humanitarian relief support.

Remote sensing has the advantage of an objective imaging device that captures data over large areas under equal conditions. This is much like a ‘neutral’ observer’s camera, just covering a much larger extent, and taken from an orbital view. Thus, remote sensing entails a trend for a ‘democratizing tool’, designed to reveal the situation ‘as is’, non-distorted, non-manipulated, and potentially accessible to everyone.

Remote sensing versus field mapping

In comparison to terrestrial field mapping and observations on the ground, EO-based humanitarian response benefits from general assets of remotely sensed data:

From a distance. The core principle of remote sensing data, its obtainment from indirect contact with the object of concern, is critical to crisis-related applications. Often the area affected by the crisis is inaccessible or difficult to reach or from a security point of view too dangerous to enter. This means that information derived from remote sensing is the only information available.

Area-wide coverage. Depending on the spatial resolution, areas can be covered with variable extent, under the same imaging conditions and characteristics. The trade-off between resolution and extent is thereby a limiting factor that is also reflected in costs and timeliness of data provision.

 Global availability. Satellite data are globally available with a theoretical cover rate of some 95% of the inhabitable space of the globe and factually 100% of the permanent settlement area. Note that cloud cover (e.g. in tropical latitudes) is a limiting factor for data acquisition and analysis.

Retrospective view. Time-series not only enable constant monitoring in future time steps, but also ex-post assessments by past sequences. This is a key factor for estimating detected trend patterns in a more reliable matter.

What kind of data to use?

There is a large variety of different optical as well as radar sensor types available today that can provide information at all kinds of spatial, temporal and spectral resolution applicable for many different humanitarian disaster scenarios.

Acquiring satellite data is a trade-off between resolution, extent, and prize.

For our applications we are using two main groups of optical satellite images:

1) Very high spatial resolution (VHR) images (pixel size <1m) are used for camp assessment and monitoring purposes

2) High resolution (HR) images (pixel size between 4 and 30m) are used for groundwater exploration and environmental monitoring

In several application scenarios the operational use of different radar satellite / SAR data techniques were successfully tested for population estimation, natural resources, and environmental impact or hazards assessments.

What do we see on VHR satellite data?

Depending on the type of camp, different dwelling structures (e.g. tents covered with white or blue tarpaulin, tukuls, makeshift shelter) can be identified besides camp infrastructure and surrounding land cover (here: Zam Zam IDP camp, Darfur, Sudan, WorldView-2).

The number of single dwellings, derived by visual interpretation or automated feature extraction, can be used for population estimations.