Satellite Earth Observation: estimating soil moisture content

Sputnik 1

Sputnik-1, planet Earth’s first artificial satellite

Earth observation satellites have been constantly flying over our heads for the past fifty years, mapping and monitoring our planet’s structure, resources and state.

The first satellite (Sputnik-1), not bigger than a beach ball, was launched in 1957 and transmitted radio signals, that could be received on Earth. These signals were used to gather information about the density of the upper layers of the atmosphere and the propagation of electromagnetic waves in the ionosphere. Since then technological advances have allowed us to put in orbit very sophisticated sensors that are able to measure the electromagnetic radiation emitted or reflected by the Earth’s surface and atmosphere with high accuracy. These measurements can be then used to extract information on many geophysical properties. In this post, and the ones that will follow, we will try to give a closer look to the different geophysical variables that it is possible to measure from space and how they can be used for practical applications. We will start the overview explaining how soil moisture is estimated from satellite measurements and how it can be used for irrigation water management.

Water is an essential resource in general but especially in areas where it is scarce and in big areas of farming activity. By definition, water management represents the use of the proper quantity of water at the proper time and it is usually pursued by combining measurements of soil moisture with an optimised irrigation plan. While the second element is easy to design if water is available, having detailed spatial information on soil moisture is still a challenge. Earth Observation data have demonstrated to be a useful source of information to this purpose.

Scientific research on the retrieval of soil moisture from satellite sensors has a long history: it began with the availability of the first satellite images. Research has been done using different sensors, spanning different parts of the measured electromagnetic spectrum, and leading to several methodologies to estimate soil moisture content. All the algorithms developed are based on the inversion of models, of analytical or empirical nature, that relate the variables measured by satellite sensors to near-surface soil moisture. Depending on the sensor employed to image the Earth’s surface, different spatial and temporal resolutions can be achieved, thus the selection of the appropriate one will depend on the type of monitoring that is planned to be done.

The Sensors

satellites

The Global Environmental Satellite Observation Network

A brief survey of remote sensing sensors as well as their suitability in soil moisture retrieval will be presented hereafter. Multispectral sensors are remote sensing instruments that can acquire data in several bands of the optical and near infra-red part of the electromagnetic spectrum. The variable that is possible to measure with this type of instrument is the spectral reflectance, i.e. the ratio of energy reflected by the Earth to incident energy, in general coming from the sun. This quantity can be directly related to surface soil moisture. A major drawback of optical instruments is their dependence on atmospheric conditions and the need of the sun as a source of illumination.

Microwave sensors acquire measurements in the frequency range from 0.3 GHz to 300 GHZ i.e., a wavelength that spans from 1 m to 1 mm. There are two main types of such instruments: passive and active. Passive microwave sensors or radiometers measure the radiation emitted by the Earth’s surface in the field of view of the instrument. Over land, the emitted radiation is mainly dependent on soil temperature and its dielectric properties, the latter being directly influenced by the soil moisture. Although radiometers are not subject to the presence of the sun and the atmosphere has a little impact on the measured signal, the resolution of such instruments is in the order of several kilometres making them useful only for studies on the global scale.

Active microwave sensors or radars send out pulses of electromagnetic radiation and measure the amount that is backscattered in the direction of the antenna sensor. Having their own source of illumination, such sensors can acquire data night and day and in presence of cloud coverage. Data acquired by active microwave radars, such as scatterometers and Synthetic Aperture Radars (SAR), are also sensitive to changes in soil moisture among other parameters. Scatterometers, with a spatial resolution of several kilometers, have been used with success for studies on the global scale, while SAR, with a resolution that can reach a few meters, is suitable for studies at the local scale.

The choice of the type of sensor has to be done depending on the application. In case of water management for irrigation, the most suitable instrument is represented by SAR given its high resolution and night/day/all weather acquisition capabilities. The soil moisture maps generated from SAR data can be used by governmental agencies that manage water distribution or by the single farmer to schedule in a more efficient way the irrigation of their fields.

A practical application for irrigation water management

In the framework of a project led by Starlab [1], the different issues concerning the  operational retrieval of soil moisture using SAR images were investigated and tested over an area located in the north of Catalonia.

 The area of study (the red dots indicate fields sampled during the survey)

The area of study (the red dots indicate fields sampled during the survey)

The approach chosen for the retrieval of soil moisture is based on the calibration of a semi-empirical algorithm [2]. The calibration attempts at the optimization of the model in order to take into account the specificities of the vegetation cover in the area of interest.

Results were compared to ground measurements acquired during the several in-situ campaigns, showing that SAR data are able to estimate surface soil moisture with an accuracy that, depending on the type of vegetation cover, can vary between 5% and 10%.

Soil moisture map generated from an Envisat ASAR image covering the Catalonian area of Alt Empordà

Soil moisture map generated from an Envisat ASAR image covering the Catalonian area of Alt Empordà

Soil moisture maps generated from SAR data can be used to monitor the evolution of soil moisture on regional scale with high resolution. This is particularly interesting for applications such as water management for irrigation where the knowledge of the fine scale distribution of the soil water content would allow to precisely detect sensible areas. When the whole of the new Sentinel satellite constellation is launched and operative, the frequency of acquisition will be between one and three days over Europe and Canada.

Here in Starlab we are offering a soil moisture monitoring service based on the combined use of SAR images and in situ probes. The figure to the right shows an example of a soil moisture map generated using an image acquired by the ESA satellite ENVISAT with the ASAR sensor.

If you’d like to know more about how we use SAR data to monitor agriculture, please head over to the vegetation health product line for more information.

[1] Reppucci, A.; Moreno, L, “Near Surface Soil Moisture Estimation Using SAR Images: A Case Study in the Mediterranean Area of Catalonia”.In Proc. ESA Living planet symposium 2010, Bergen (Norway)
[2] : Loew, A., Ludwig, R.; Mauser, W. (2006). Derivation of surface soil moisture from ENVISAT ASAR Wide Swath and Image Mode data in agricultural areas. IEEE Trans. Geosci. Remote Sens., vol. 44, pp. 889-899.
Bookmark the permalink.

Leave a Reply

Your email address will not be published. Required fields are marked *