Satellite Imaging – A New (Information) Wellspring


An Ongoing Paradigm Shift And The Road Ahead

Deepening of knowledge of the factors influencing the environment, particularly those related to the development of plants, which can be monitored and analyzed over their various stages of growth from space, is foreseen as one of the main application areas for new satellite missions.

To take advantage of the research and new breakthroughs in the field of photogrammetry and space technologies, it was realized that there is a need for altering formerly used data policies and liberalizing data access in order to foster development of new applications and proliferate user base. Such trend has been observed globally and brought new opportunities for improving the way satellite imaging is perceived and how it impacts our everyday life.

Increased availability of high resolution imaging has brought significantly improved level of detailing. Streets and railways, individual buildings, urban layouts, infrastructure, and related activities can be precisely mapped and detected, as well as individual agricultural fields and crop structure.

Crop classification map derived from satellite multispectral time series. Resolution enhancement is achieved through the process of fusion of measurements obtained from different instruments by application of a specially designed data fusion algorithm that takes the most information from each imaging instruments (lower quality maps corresponding to standard approaches are shown on top).

Land use / land cover maps derived from satellite measurement have been traditionally seen as valuable sources of land surface information on different thematic levels. With availability of new imaging instruments and improved processing techniques, quality of thematic maps became suitable for development of various, novel user-oriented mapping services dedicated to agriculture, forestry, spatial planning and risk diagnostics.

The range of operational instruments with different imaging capabilities that are becoming available on board of different satellite platforms at the time is constantly growing, with an increasing rate. It is also followed by the development of new facilities and information infrastructure on the ground, which opens space for new business opportunities and further advancements in the related R&D fields, including sustainable production of safe and healthy food through application of new technologies, which is one of the challenges in the focus of Institute’s work

satellite-1

Satellite detection of NDVI (Normalized Difference Vegetation Index) is used to estimate crop condition throughout the season.

Although satellite products at sub-meter and meter spatial resolution are still only commercially available, there are many imaging instruments with free data policy that are suitable for continuous monitoring of plants’ bio-physical parameters and applications involving production of different thematic maps, on a daily or weekly basis. This is particularly true in the case of the first of the two Sentinel-2 satellites, which is part of the Copernicus-The European Earth Observation Programme, and which became operational in November 2015. It already provides free, high quality, multispectral data at spatial resolution of 10 m every 10 days, and after planned launch of the second satellite the Sentinel-2 mission will provide images of the same point on Earth every 5 days.

Copernicus

“Copernicus addresses both the acquisition of Earth Observation data AND the provision of value-added information through the Copernicus services for land, marine, atmosphere and climate change monitoring as well as for the management of emergency and civil security.” – Mr. Philippe Brunet, Director of the Directorate-General for the Internal Market, Industry, Entrepreneurship and SMEs, European Commission

Important applications involving production of thematic maps include assessment of human induced or natural risks (like flooding and droughts) on yield estimation and plant health status in general, but also activities related to efficient land management on different levels and creation of adequate, accompanying agricultural policies.

Considering that some of the common characteristics of materials and devices developed for applications in space are usually high reliability and availability, the need for adequate handling of uncertainty in applications of satellite imaging in agriculture and other food production fields arises from the inheritable complexity and variability of the processes of interest that are influencing or characterizing plants development in the field. Resolving this type of uncertainty is something the Institute is oriented towards and makes effort in different ways, e.g. through design of new machine learning and remote sensing techniques, novel data platforms and services, or new sensor devices for use in the field.

Color composite

NDVI index computed from red and NIR band

Novi Sad and river Danube

False color composite consisting of near-infrared (NIR), green and blue band

False color composite consisting of blue, NIR and SWIR2 band

Novi Sad and the river Danube, Serbia, 30 m Landsat 8 / taken on the 21st of July 2015.

Growing ecosystem of different optical and radar based imaging instruments offers enough diversity for managing different application determined requirements, such are appropriate spatial, spectral or temporal resolutions of acquired space-borne measurements or on the other hand all-weather imaging capabilities, but also brings many additional challenges that need to be addressed properly. This includes handling of massive amounts of data on previously unprecedented scale (regarding their volume due to high spatial and temporal resolutions, their variety as a consequence of different imaging techniques, or their velocity in
the case of real time applications), but also the need for the design and development of efficient algorithms and analysis tools – customized for specific domains and fields of applications. It is estimated that the fully operational Sentinel-1,-2,-3 missions will be jointly producing around 8 terabytes per day, which does not include user generated data.

“At over 40 years, the Landsat series of satellites provides the longest temporal record of space-based surface observations.” – Dr. David Roy, professor at the Geospatial Sciences Center of Excellence, South Dakota State University, and the co-chair of the USGS NASA Landsat Science Team

Therefore, it can be said that besides the fundamental change produced by the new data policies in the field of satellite imaging and remote sensing we are currently witnessing the paradigm shift that can be also observed in other fields that involve managing of large volumes of data.

The synergistic effect of combining satellite imaging and task specific auxiliary data (e.g. data collected from the field) will enable better understanding and control of the processes that are responsible for ensuring food security and its sustainable production.