Report in reference to Trends in Sustainable Energy

Available online at www.sciencedirect.com

www.elsevier.com/locate/asr

ScienceDirect

Advances in Space Research 57 (2016) 110–126

Advancements in medium and high resolution Earth observation for land-surface imaging: Evolutions, future trends and contributions

to sustainable development

Yashon O. Ouma

Space Technology and Geoinformatics, Department of Civil and Structural Engineering, School of Engineering, Moi University, Eldoret, Kenya

Received 25 April 2015; received in revised form 24 October 2015; accepted 26 October 2015 Available online 31 October 2015

Abstract

Technologies for imaging the surface of the Earth, through satellite based Earth observations (EO) have enormously evolved over the past 50 years. The trends are likely to evolve further as the user community increases and their awareness and demands for EO data also increases. In this review paper, a development trend on EO imaging systems is presented with the objective of deriving the evolving pat- terns for the EO user community. From the review and analysis of medium-to-high resolution EO-based land-surface sensor missions, it is observed that there is a predictive pattern in the EO evolution trends such that every 10–15 years, more sophisticated EO imaging systems with application specific capabilities are seen to emerge. Such new systems, as determined in this review, are likely to comprise of agile and small payload-mass EO land surface imaging satellites with the ability for high velocity data transmission and huge volumes of spatial, spectral, temporal and radiometric resolution data. This availability of data will magnify the phenomenon of ‘‘Big Data” in Earth observation. Because of the ‘‘Big Data” issue, new computing and processing platforms such as telegeoprocessing and grid- computing are expected to be incorporated in EO data processing and distribution networks. In general, it is observed that the demand for EO is growing exponentially as the application and cost-benefits are being recognized in support of resource management. � 2015 COSPAR. Published by Elsevier Ltd. All rights reserved.

Keywords: Earth observation (EO) trends; Sustainable development; EO Big Data; Grid-computing; Telegeoprocessing; Value-added geoinformation

1. Introduction

For the past 50 year, satellite based observations have provided effective ways of monitoring the planet, thus aid- ing in the exploitation and management of Earth’s resources. In essence, Earth observation (EO) data and data products are significant in making of reliable and informed monitoring and management decisions in differ- ent areas such as: agriculture sustainability, climate change, Earth system science, human health and epidemiology, energy management, biodiversity and ecosystem studies and water resources management (GEO 2009).

http://dx.doi.org/10.1016/j.asr.2015.10.038

0273-1177/� 2015 COSPAR. Published by Elsevier Ltd. All rights reserved.

E-mail address: [email protected]

From global to local footprints, geoinformation derived from Earth observations are now significantly contributing towards the study of Earth systems. Understanding the interactions and variations of Earth systems – comprising of the geosphere, biosphere, hydrosphere, cryosphere and the atmosphere, is quite critical in protecting the global environment and in return enhancing the wellbeing of human health, safety and welfare. It is now agreeable that Earth observation technology from space is most suitable for providing a global, repetitive and continuous data cov- erage for the Earth’s surface and for understanding the Earth system. In brief, EO data products are now being used for the effective study and analysis of short-to-long- term interactions, changes and variations on the planet as depicted in Fig. 1.

Y.O. Ouma / Advances in Space Research 57 (2016) 110–126 111

Earth observation in general comprises of the measuring technologies and platforms, and refers to the passive and active observation or remote sensing of the Earth from space borne sensors. A continuous review and analysis of the availability of Earth observation data, the societal ben- efits (e.g. Fritz et al., 2008), and the inference of the future perspectives in terms of data types and modes of acquisi- tion and accessibility is important to the EO user commu- nity especially in Earth systems representation and modeling. In addition to active and passive EO data, Earth observation supporting systems such as telecommunication data relaying satellites and Global Navigation Satellite Systems (GNSS) continue to contribute to the broader perspective of Earth observation and geoinformation.

Since the launch of SPUTNIK-1 in 1957 as the first Earth artificial Earth orbiting satellite for gravity field measurements, which was followed by Earth observation reconnaissance CORONA satellite in 1959 (McDonald, 1995), Earth observation satellites have continued to mon- itor the global environment. Data from EO satellites have been used to effectively demonstrate the inherent fragility of the Earth system and the exposure of the system to rapidly growing human-induced stresses.

In terms of the application of EO, it can be recognized that much has not only changed in terms of the mainstay application of satellite-based military and defense surveil- lance, but also when it comes to EO applications in civilian use (private industry, governments and academia), a lot has since changed with regards to the acquisition and process- ing techniques, applications, and in general the community user base has equally increased. In light of the developments in EO, this paper presents an analytical overview on the

Fig. 1. Components of and interactions within the Earth system. (I

socio-economic benefits or societal benefits of EO to sus- tainable development, that has stimulated the tremendous interest and growth in Earth observation, and looks at the different eras of satellite remote sensing in terms of sensor resolutions and applications for land-surface imaging, with the view of inferring the future imaging system demands and data access and processing requirements.

Despite advancements in EO technologies, major hur- dles on issues to do with awareness and system maturity to capacity building, investment levels and cultural barriers between the supplier and user communities still remains (Tatem et al., 2008). The remainder of this paper is orga- nized as follows. Section 2 highlights the benefits of Earth observation and motivates on the need for providing value- added and intelligent EO based geoinformation in support of sustainable development. Section 3 provides an overview on the evolutions, the future trends in Earth observation systems for medium and high-resolution land-surface imag- ing, the EO system mass-payloads and the emerging issue of EO Big Data. Section 4 presents an outlook on the future of EO data acquisition design capabilities and advancements in data access and processing technology and Section 5 is finally on the conclusions and recommendations.

2. Earth observation for geoinformation: opportunities and

challenges

2.1. Benefits of Earth observation

The synoptic view through satellite imagery offers the most appropriate method for quick, economical and

mage Courtesy of NASA: Geophysical Dynamics Laboratory).

112 Y.O. Ouma / Advances in Space Research 57 (2016) 110–126

reliable mapping and monitoring of Earth’s resources, and for change detection through repetitive satellite remote sensing over various temporal and spatial scales provides. The repeatability of Earth observations thus permits the observation of extensive areas and a number of environ- mental indicators and therefore serves an important basis for decision making.

The availability of remotely sensed data from different sensors on various platforms, with a wide range of spa- tial–temporal, radiometric and spectral resolutions has made remote sensing the best source of data for different scales from local to global (Melesse et al., 2007). The con- tribution of EO towards monitoring of the Earth system is attributed to the direct and or indirect Earth observations, measurements and products (geoinformation) which con- tribute to the construction of policies that are used in sup- porting Earth systems management for sustainability.

Notably, EO products have changed the paradigm of mapping and modeling, by sensing beyond the visible spec- trum at improved spatial resolutions. Such data have enhanced the ability of understanding and measuring: rain- fall, surface and groundwater, weather patterns, human and vegetation health, atmospheric pollutants, soil mois- ture and rock types, amongst other applications that com- prise Earth system studies. Different Earth system models are now able to incorporate EO based geoinformation with the objectives of not only understanding the present, but also the future scenarios and phenomenon of the Earth sys- tem for sustainable management and in support of devel- opment. A summary of the broad areas of applications of Earth observations is presented in Table 1.

Because of the increased contributions and benefits of EO in sustainable resources management and development programs, the 1992 UN conference on Environment and Development specifically endorsed the use of Earth obser- vation for sustainable development. The conference declared that the guiding principle for human activities in the future and the Earth’s limited natural resources, need to be carefully managed together with the natural environ- ment by using actionable data through Earth observation. Other related global endorsements on the role of EO in sus- tainable development include: (a) the 2002 Johannesburg World Summit on Sustainable Development (WSSD) on using Earth observation technologies; (b) the 2007 Cape

Table 1 Applications and benefits of Earth observation products (Borzacchiello and C

Area of EO data application Application ob

Disasters Reducing loss Health Understanding Energy Improving ma Climate Understanding Water Improving wa Weather Improving wea Ecosystems Improving the Agriculture Supporting su Biodiversity Understanding

Town Declaration by Group on Earth observations (GEO) – Ministerial Summit on Earth observations for Sustainable Development (EOSD), and (c) the 2012 Rio +20 UN Conference focusing on Sustainable Development using Earth observations.

Fig. 2 presents a summary of some of the specific appli- cation areas of EO data with spatial resolutions ranging from medium- to high spatial resolutions. The application areas presented in Table 1 and in Fig. 2 can be categorized into the following interrelated areas of applications: (i) long term satellite data records; (ii) assessment of land sur- face parameters and processes, e.g. ecosystem net primary productivity and evapotranspiration; (iii) monitoring of flooding, landslide and permafrost-related infrastructure; (iv) supporting environmental data for urban map updat- ing and high-resolution digital elevation models.

In all the applications, the significance of any EO product relies on the value of the information used. This implies that the EO data processing phases are significant in the conver- sion of Earth observation data into reliable geoinformation and knowledge mining, which are necessary in the decision making and policy formulation. The phases in the EO data mining are schematically represented in Fig. 3. The phases are bottom-up based, with the decision support being the final goal. This implies that failures in decisions based on EO products are mostly attributed to failures in EO data processing and accuracy. From Fig. 3, it is factual that data in itself is of very little value and benefit unless processed properly and purposefully. The value-added EO products and services are the most significant in making the final deci- sion at different levels of policy formulation.

Despite the availability of EO data coupled with the powerful data processing platforms, and the ancillary EO data sources such as GNSS, the incorporation of Earth observation products into routine resource management operations continues to be challenging. These challenges can be attributed in part to the following contributing fac- tors: first, there is lack of sufficient funds due in part to budget cuts, such that some of planned EO based activities are often stagnated or postponed. Secondly, there are what can be termed as ‘‘communications and cultural barriers” between the space scientists and data users that are often difficult to overcome. These challenges can generally be viewed as economic, technical and cognitive in nature.

raglia, 2011; GEO, 2009).

jective and benefit description

of life and property from natural and human-induced disasters environmental factors affecting human health and well-being

nagement of energy resources , predicting, mitigating and adapting to climate variability and change ter resource management through better understanding of the water cycle ther information, forecasting and warning management and protection of terrestrial, coastal and marine resources stainable agriculture and combating desertification , monitoring, and conserving biodiversity

Fig. 2. Role and applications of EO-based geoinformation in decision-making for sustainable development (Euroconsult, 2012 and Révillon, 2012).

Data Processing: Geoinformation data mining and products

Knowledge transformation for specific application

Decision Support

and Policy Formulation

Earth observation data acquisition

EO data value addition process

for intelligent geoinformation

products

Fig. 3. The processing chain for EO-based geoinformation products in support of decision making and policy formulation for sustainable resource management and development.

Y.O. Ouma / Advances in Space Research 57 (2016) 110–126 113

While there are attempts to improve on the technical barriers such as having access to data, constraints related bandwidths for data download and upload, storage, lim- ited tools for visualizing and techniques for integrating these data into the envisaged models still impair the uptake, analysis and representation of EO data. Other related impediments include:

(1) managing exponentially growing data volumes; (2) appropriately describing EO metadata to ensure

long-term utility and re-usability; (3) providing data in formats that enable fast and univer-

sal integration; (4) providing clear and easy discovery of and access to

EO data and information products;

(5) collecting data at risk to extend the environmental data record, and

(6) provision of precise, accurate quality geodata.

If these drawbacks can be harnessed, then there will be actionable data to a majority of interested users and in addition there will be value-added EO-based geoinforma- tion products for the desired socio-economic benefits in terms of sustainable development.

In order to overcome the above challenges and draw- backs, different initiatives ranging from EO policies, data access and processing should be considered for value- addition to EO products. Such initiatives include the Group on Earth Observation (GEO) which is a leading a worldwide effort to build a Global Earth Observation

114 Y.O. Ouma / Advances in Space Research 57 (2016) 110–126

System of Systems (GEOSS) policy for EO data access. In addition to GEO, the Open Geospatial Consortium is working on EO data standards. Towards access and pro- cessing of EO data, cloud computing and on-demand grid processing platforms are gaining significance in coping with the large volumes of EO data.

2.2. Value-addition to EO data

Since the civil community started using public EO data, most of the effort has been geared towards image classifica- tion for thematic mapping, with the main objective of creating inventories about land use and land cover (Wilkinson, 2005). Besides classification, land use and land cover change detection (thematic mapping), EO data has also been used in the spatial analysis and modeling of varied Earth system phenomena as shown in Fig. 4. Nonetheless, the methodologies used for thematic mapping and or change detection have not grown as exponential as the growth in EO data spatial, spectral and temporal resolutions (Coppin et al., 2004).

Besides the geographical based thematic mapping and spatial modeling, new frontiers of EO-based applications are emerging. Such areas include more technical and inter- disciplinary applications such in Transportation Systems and Structural Engineering, where nondestructive based structural health and functional monitoring and evalua- tions of engineering structures can now be carried out on a routine basis in real-time (Ahlborn, 2011). It is with this backdrop that value addition to original EO image pixels should as much as possible represent and depict the real-world objects, i.e. intelligent geoinformation (Zhou, 2002). Such a representation of image features and

Map of categorical Variables

Map of thematic classes

Thematic Remote Sensing

Image Classification

Fig. 4. Typical applications of EO data in class

extraction can only be achieved by analyzing pixels as image objects (Hay and Castilla, 2008).

According to Chen et al. (2012), in order to add-value to EO-based geoinformation, there should also be focus on the conversion of image pixels-to-objects, so as to more accurately represent the real-world. Such an approach will be instrumental in creating intelligent geoinformation, through accurate representation of the real-world objects as seen by the human visual-based image cognition system. Such approaches could include processing based on support vector machines (Tzotsos et al., 2011); genetic programming (Dos Santos et al., 2010); neural networks (Del Frate, 2007); wavelets transform (Ouma et al., 2006; Ouma et al., 2008).

Increasingly therefore, limitations in satellite-data appli- cations are slowly shifting from the technology of acquiring the data to the techniques required to optimally exploit the available data for value-added geoinformation. Further- more, with the vast range of passive and active based EO data to choose from, improvements in data processing and multisensory data fusion could help eliminate cloud- obscured and nighttime data loss, and provide multi- image virtual databases for modeling of environmental and social processes.

3. Trends in satellite based imaging systems for Earth

observation

3.1. The eras in Earth observation systems

According to Tatem et al. (2008), the more than half a century of satellite based imaging has provided both iconic views and unprecedented scientific insights into Earth

Map of continuous variables

Quantitative Remote Sensing

Modeling and Spatial Analysis

ification and spatial analysis and modeling.

Y.O. Ouma / Advances in Space Research 57 (2016) 110–126 115

observations. Such insights have been realized in, but not limited to, the following field: (i) observation of the Earth through remote sensing satellites and meteorological satel- lites; (ii) specific missions for astronomy, geodesy and observations of the sun, ionosphere, high atmosphere, meteorites and magnetosphere; (iv) satellites for GNSS, and (v) special topics based satellites such for radioactive measurements and surface and groundwater mapping. These areas of development have equally lead to the launch of EO satellites with different resolutions, and often tai- lored to specific applications.

Fig. 5 shows growths that have been experienced in the launch of civilian EO imaging satellite systems over the past nearly 50 years of EO activities. The number of launched satellites and the spatial-spectral resolutions are respectively illustrated in Fig. 5(a) and (b). Fig. 5(a) also presents the number of expected EO satellite launches in 2015, with specific reference to optical-based passive imaging sensors.

(

(

Fig. 5. (a) A survey of optical-passive satellite-based EO activities in terms of t passive based Earth observation sensors with different spectral resolutions.

Within the timeframes presented in Fig. 5(a), it is observed that activities in satellite-based Earth observation gained momentum in the late 90s. The late 90s also marked the realization of the high spatial resolution (HSR) and very-high spatial resolution (VHR) Earth observation data acquisition from satellite borne sensors, which were previ- ously dominated by airborne photogrammetric imaging. With the recorded increase in EO launches from 2010 to 2014, it can be projected that at least fourteen EO imaging satellites will be launched in 2015 (Révillon, 2012).

As far as advances in EO spectral resolutions are con- cerned, Fig. 5(b) illustrates the rapid emergence and prolif- eration of panchromatic, thermal and hyperspectral EO data. This is observation is in concurrence with the increased EO launch activities depicted in Fig. 5(a). It is observed that from the late 1990s, the panchromatic data- sets were acquired in higher spatial resolutions and so did the spectral resolutions increase from multispectral to hyperspectral sensors. This trend can not only be attributed

a)

b)

he number and type of satellite launches. (b) Spatial resolution of optical-

116 Y.O. Ouma / Advances in Space Research 57 (2016) 110–126

to the growing EO user demands, but also to the techno- logical advancements in the satellite sensor and imaging design techniques. Within the same timeframe, as in Fig. 5(a), the multispectral sensors have also improved in the spatial resolutions, as evident in Figs. 5(b).

From the design and launch viewpoint, it observed that from the early 1990s, two distinct trends in satellite design and operation have emerged. First, the large national and international space organizations, e.g. NASA and the European Space Agency (ESA) respectively, focused their EO resources on the design and launch of large multisen- sory platforms, such that each sensor was designed to mon- itor a specific aspect of Earth system processes, frequently at the global scale. Examples of such imaging systems include the Terra (EOS AM) and Aqua (EOS PM), which were the first of a series of multi-instrument spacecraft forming NASA’s Earth Observing System (EOS). This was followed by the National Polar Orbiting Environmen- tal Satellite System (NPOESS), in 2013. Parallel to NASA’s activities, 2002 saw the launch of ESA’s Environmental Satellite (ENVISAT), which carries 10 different sensors at the size of a double-decker bus. ENVISAT was the largest EO space-based imaging system ever built (Tatem et al., 2008).

The second but contrasting trend in satellite design is on the development of not only large, but also smaller and consequently cheaper satellites. The small satellite concept, in which more than 20 countries, developed and developing are now either developing or operating remote sensing satellites at the national governmental levels or at intergov- ernmental levels (Tatem et al., 2008). This is because of the modernization and miniaturization instrument design and size, and the resulting lowered launching costs. Notably, most of these new satellites are developed and launched by either commercial operators and or in collaboration with the national governments (Murtaza, 2011). The trends in the mission sizes and objectives are summarized in Table 2 (Zhou et al., 2002), and discussed in Melesse et al. (2007).

Following the observed trends and advances in EO and imaging trends depicted in Fig. 5 and Table 2, it can be argued that apart from panchromatic, multispectral and thermal imaging, the era of satellite based hyperspectral imaging has reached with the example of EO-1 Hyperion. In addition to passive optical sensors, digital elevation models (DEM) are now available from synthetic aperture radar such as the Shuttle Radar Topography Mission (SRTM), TerraSAR-X and TanDEM-X at varied spatial resolutions. Parallel to the advances in hyperspectral and SAR imaging growths, there is also the emergence of con- stellations and domain specific EO platforms such as the Disaster Monitoring Constellation (DMC); European Glo- bal Monitoring of Environment and Security (GMES), now called Copernicus programme, following from the Terra-Aqua and ENVISAT satellites, with lower spatial resolutions but higher multispectral resolutions and increased frequency of coverage, which can now effectively

monitor events which affect the environment such as fire, flood and desertification. These developments are likely to herald the post-2010 era of Earth observation (Wiscombe et al., 2002).

An example of the post-2010 Earth observation satellites is the ESA-Copernicus programme on the Sentinel satellite missions, which comprises of the most comprehensive Earth observation system worldwide for RADAR and optical imaging, which will be useful for a wide array of applications ranging from ocean and global land monitor- ing, atmospheric composition monitoring and for high pre- cision ocean altimetry measurements (Malenovsky et al., 2012). The Sentinel-1 for example, can now be used for the production of high resolution data products, depending on the acquisition mode(s). Similarly as of July 2015, pre- liminary data from Sentinel-2A multispectral imager is now available for systematic acquisition in a single obser- vation mode. In terms of the applications, the Sentinel-2 with 10–20 m spatial resolution, 5-day revisit frequency, global coverage and compatibility with to the Landsat mis- sions offer new opportunities for regional to global resource monitoring. The potential application areas include: land monitoring, emergency management, security and climate change, and associated thematic applications.

The increasing benefits from the applications of EO data have resulted into the demand for more passive and active based imaging data with improved spectral and spatial res- olutions (Révillon, 2012). As derived from the EO develop- ments in Table 2, it is evident that today the most widely used EO products irrespective of the spectral resolutions and whether passive or active sensor, have a spatial resolu- tion ranging from 2.6 m–30 m. Thus the near recent launch patterns in Fig. 6 are largely influenced not only by the technological advancements in the imaging techniques and the costs related thereto, but also by the specific data application demands such as in disaster management, mar- itime monitoring and infrastructure management.

With the exponential trends as depicted in Fig. 6 it is projected that there will be about 288 optical and SAR based Earth observation satellites from 2007 to 2017, for both civil, commercial and defense purposes (Révillon, 2012; Euroconsult, 2012). In addition to the patterns in Fig. 6, it is observed that the design features of the imaging instruments are being customized around the user and application demands. These variations are typified by the minimum size of objects distinguishable on the Earth’s sur- face (ground spatial distance); size of the region of the elec- tromagnetic spectrum sensed (spectral resolution); number of digital levels used to represent the data collected (radio- metric resolution) and the intervals between imagery acqui- sition (temporal resolution).

In general, it can be argued that the increasing number of EO satellites and the increasing availability of raw data will lower data costs. This is true for unprocessed data, however for processed data; the costs are very still very high. Furthermore, as the demands for user specific data acquisition increases, there is no doubt that there will be

Table 2 Timeline on the evolutions and developments in Earth observation systems and sensors (Zhou et al., 2002; Melesse et al., 2007).

Timeline and era of Earth observation systems and sensors Earth observation development and image acquisition

Napoleonic times: Hot Air Balloons and Airborne remote sensing

The airborne remote sensing era evolved during the first and the Second World War (Avery and Berlin, 1992). During this time remote sensing was mainly used for the purposes of reconnaissance, surveying and mapping, and military surveillance (Melesse et al., 2007)

1957–1970s: First generation of Earth Observation Systems (EOS-1)

The spaceborne EO era began with the trial-and-error launch test-concept based imaging satellites. EXPLORER-1, CORONA, ARGON and LANYARD as the first operational imaging satellite reconnaissance systems (McDonald, 1995). Between 1960–1972 these sensors mostly acquired panchromatic and in some cases color and stereo classified data (Zhou and Jezek, 2002). These EO platforms were followed by the first meteorological satellite – Television and Infrared Observational Satellite-1 (TIROS-1)

1972–1986: Second generation – civilian optical based Earth Observation (EOS-2)

The launch of Landsat 1 in 1972 symbolized the modern era of civilian Earth remote sensing, for Earth resource mapping (Lillesand and Kiefer, 2000). The acquired image data was directly delivered in digital format for the first time and for civilian utility

1986–1997: Third generation of Earth Observation era and emergence of active microwave imaging (EOS-3)

In what is considered as the third generation, EO experienced significant development in technologies and demand driven applications, which was characterized by the launch of SPOT-1 in 1986, carrying two High Resolution (HR) Visible sensors, With its 10 m panchromatic resolution band SPOT-1 was the first EO satellite capable of stereoscopic imagery in crosstrack. Synthetic aperture radar (SAR) sensors such as ERS-1, JERS-1 and ERS-2 (Zhou et al., 2002)

1997–2010: Fourth generation private industry comprising of high-resolution and hyperspectral EO age (EOS-4)

In this era, the EarlyBird in 1997 provided a 3 m of ground sample distance in panchromatic mode, and opened the new era of VHR EO satellites (Zhou, 2001). This had potentially applications in mapping of rivers, buildings and linear civil engineering structures. IKONOS and QuickBird followed suite in delivering VHR imagery, which is widely accepted for most GIS based mapping at scales comparable to airborne photogrammetry (Stoney, 1996)

Y.O. Ouma / Advances in Space Research 57 (2016) 110–126 117

cost implications. This will be such that the higher the spa- tial resolution, the higher the cost per square kilometer or even square meter of data becomes. In addition, factors such as the number of regions of the spectrum for which data are collected, the time taken to revisit the same area of the Earth, the spatial extent of images produced and whether the satellite’s orbit follows the sun-illuminated sec- tion of the Earth (sun-synchronous) or remains over a fixed point on the Earth (geostationary), will also influence the EO data costs. That is, the high altitude geostationary orbits result into coarse geometric resolution data, which are lower in cost per square kilometer as compared to the lower altitude sun-synchronous orbiting satellites or low- Earth orbiting (LEO) satellites.

3.2. Payload-mass trends for Earth observation imaging

systems

There is an emerging and growing trend in in the size and performance of the EO sensors towards small satellites (Xue et al., 2008; Wicks et al., 2000; Rast et al., 1999). Arguably, a small-payload would imply that the electronics and detection sensor must be reduced in the sizes and mass. The evolution of interferometers and spectrometers is a good example with respect of the miniaturization of sen- sors. In the shift from large to small satellites, there must be the balancing effect between the demand for increased performance and the technical solutions towards the satis- faction of lighter and smaller instrument payloads. With respect to active sensor, such mass-payload requirements

will dictate the size of the overall mass budget for the mis- sion, as compared to passive instrumentation (Sandau et al., 2010).

The general advantage of small satellites is that they provide capabilities which larger systems may not match, e.g. small satellites can potentially operate from any orbit thus providing greatly enhanced revisits to orbit and thus much higher temporal resolution (Fig. 7). Thus with the observational networks at lower orbital inclinations of small satellites, shorter revisit cycles with larger coverage will be possible, from the small, modular and resource- limited satellites for EO applications (Eves, 2011).

Small satellites are able to produce higher quality ima- gery due to the fact that their agility allows them to collect more light from a given scene. Agility here means that the small satellites are able to achieve attitude-change maneu- vers much faster that larger platforms (Peri et al., 2001). Such small-agile satellites are designed for a number of sci- entific applications and from different user groups (Neeck, 2015). Within the broad applications, examples of small satellites includes: the DMC small satellites constellation, RapidEye, SARAL, CALIPSO, GOCE, JASON, ADM- AEOLUS.

Notably, the very first Earth observation satellite to be launched was the CORONA satellite, which served as a classified reconnaissance satellite for mapping and spying, with a high spatial resolution of between 2 and 3 m (NRO, 2015; Jacobsen, 2005). Could this imply that the future in satellite imaging technology in terms of mass- payload is reinventing itself? The answer is definitely yes,

(a)

(b) Fig. 6. Patterns in spatial resolutions of: (a) optical-passive and (b) active- SAR based EO satellites.

Fig. 7. Orbital inclinations and the temporal agility of small and modular E

118 Y.O. Ouma / Advances in Space Research 57 (2016) 110–126

since more and more space organizations and groups are working at the development of Earth observation satellites with higher mass-payloads (Smith et al., 2010).

In summary, when looking into the future, small space- craft mass, low space segment cost and shorter develop- ment time, will create low barriers into market entry and therefore the cost-to-time ratio will enable the emergence of more and newer EO missions. On the hand, significant challenges in managing, processing and interpreting the rapidly acquired and voluminous EO data will result into the ‘‘Big Data” phenomenon.

3.3. The EO ‘‘Big Data” phenomenon

With the exponential growth of data amounts from EO systems, along with the increasing degree of diversity and complexity, EO data is soon being faced with the geo- database problem of ‘‘Big Data”. This will be enhanced by the numerous sources of EO data that are available and the amount of data in terms of resolutions and size. As defined by Che and Zhang (2014), Big Data phe- nomenon or problem is experienced when a large collation of data sets whose volume at any given time and scale sur- passes the state-of-the-art systems for archiving or storing and processing the data.

With regards to EO data, ‘‘Big Data” will not only refer to the volume and velocity of data versus the storage and computing capacity, but it will also refer to the variety and complexity of the remotely sensed data. The Big Data phenomenon in EO data is likely to be complicated and enhanced by the advent of the high-resolution EO systems, which will automatically lead to high dimensionality and volume of the imagery from the different EO data plat- forms such as terrestrial, aerial, near-space and far-space.

O satellites illustrating the high global revisit frequencies (Eves, 2011).

Y.O. Ouma / Advances in Space Research 57 (2016) 110–126 119

The Big Data phenomenon in EO can be justified cur- rently by the fact that from orbital space platform alone, there are more than already 200 EO satellite sensors cap- turing multispectral, multispatial and multiemporal optical and SAR based EO data (Esfandiari et al., 2006; Révillon, 2012). Some of the data comprise of the high resolution sensors onboard SPOT-7 and GeoEye-1 (passive) and ALOS-2 (active) with respective spatial resolutions increasing to less than 1 m and 0.5 m, and the medium spatial resolution sensors like the Landsat-8 OLI with panchromatic spatial resolutions of 15 m. And also lower resolution sensors such as NOAA-AVHRR and MODIS also continuing to acquire data at more than 200 m.

Similarly, from the spectral resolution standpoint, hyper- spectral imagers such as Hyperion instrument and HJ-1A satellite with a hyperspectral sensor are acquiring spectral information with more than 120 bands at medium resolu- tions. Following this trend, EO with varying and improved spectral, spatial, radiometric and spatial resolutions will inevitably result into high volumes of EO data availability and complexity in terms of data handling and processing.

Despite the expected complex multidimensionality and size of EO data in future, it is also arguable that the emer- gence of ‘‘EO Big Data” will contribute to the increased availability of remote sensing data to the user community. As depicted in Fig. 8, the future data dimension is expected to increase such that the future dimension (p � p) will be much larger than that of (n � n) in terms of the spatial- spectral-temporal and radiometric data dimensions.

With the ‘‘Big Data” in EO being a reality in the very near future, and coupled with the demand for real- and near real-time processing capability for many time-critical applications, the issue of complex-data-intensive manage- ment and processing will definitely become a priority in EO systems. This implies that advances in EO systems can- not be discussed now without discussing the frontiers in the processing platforms for future EO data.

4. Future Earth observation advances and data accessibility

With the increased volume of remote sensing data, there are the already mentioned challenges in the processing, publishing, discovery, and exploitation of EO data. In res- onance to the developments in EO data acquisition, the computing platforms have to also improve in order to keep up with the data volumes and production urgency. How- ever, it is also arguable that the ability of users to access, especially high-resolution data, far exceeds the capacity to analyze the data automatically. As such, techniques for automated production of geoinformation from EO as well as assisted image analysis are urgently in demand.

4.1. Potential advances in Earth observation: trends and

future observation systems

While the role and demand for Earth observation data continues to rise, developers and users are faced with new

opportunities and challenges that include higher specifica- tion requirements. This, as already been stated, has resulted into the evolution of EO sensor. From the past analysis, a timeline trend in the advances in EO technology presented in Fig. 9. The trend line in Fig. 9 answers the question which was posted earlier, as to whether there is a reinvention on the EO data acquisition principles, with respect to the spatial and spectral resolution and on the size of the satellites. While Fig. 9 represents the trends in the spatial resolutions optical sensing systems, the can be said about SAR imaging sensors, which have also exponentially reduced their spatial resolution (Révillon, 2012).

In Fig. 9, the emerging and newer multispectral imaging satellites are increasingly becoming available at higher ground sample distances of less than 1 m. Similarly the spa- tial resolutions of hyperspectral, thermal and SAR based sensors are also improving. This new comprises of the emergence of VHR sensors typified by IKONOS and Quickbird satellites and their revolutionary means of EO data collection. The latter is from RapidEye’s satellite con- stellation of 5 satellites with near-daily coverage of any spot on Earth at 6.5 m spatial resolution with 5-spectral bands. This growth is characterized by the emergence of small satellites which are designed for specific applications such DMC, Copernicus programme and other satellites like TOPSAT and NigeriaSat-2 which are capable provid- ing essential information on ocean, ice, land environments, and the atmosphere (Sweeting et al., 2012).

As seen in Fig. 9, there tends to be a near-sinusoidal trend on the resolutions of EO imaging systems over time. The characterization of the trend for the next generation of sensors can be inferred from the planned lunches such as ALOS-3 (2015), Sentinel-2B (2016), RADARSAT Constel- lation Mission (RCM) (2018), KOMPSAT-2B (2019), Tan- DEM (2020). The characteristics of most of the future EO systems are such that they are not only small-payload, but they are also designed for specific resources mapping. As an example, the ALOS-3 is capable of providing: high- resolution imagery with a GSD of less than 1 m on a swath of 50 km; pan-sharpened imagery with simultaneous acqui- sition of panchromatic band and four-color multispectral imagery; stereo imagery from different view angles, and enables the capability of image acquisition of any point in a wide area and with pointing to achieve timely monitor- ing if a disaster occurs.

On the front of the development of hyperspectral sen- sors, up to 220-band imaging sensors with 30-meters GSD has been realized. This is a great improvement form the initially very coarse spatial resolutions of about 250 m. The success of space-borne hyperspectral sensing technology is demonstrated by initiatives such as the Envi- ronmental Mapping Program (EnMAP), the Hyperspectral Infra-Red Imager (HyspIRI) from NASA and the Hyper- spectral Precursor of the Application Mission (PRISMA) from ASI (Staenz and Held, 2012).

On the development of the SAR imaging systems, it is likely that the future generation SAR imaging sensors will

Fig. 8. The concept of EO data acquisition and evolution of ‘‘Big Data” in terms of data dimensionality, volume and rate of acquisition.

Fig. 9. Average temporal trends in the evolution and advances in optical based sensing.

120 Y.O. Ouma / Advances in Space Research 57 (2016) 110–126

be towards the design of constellations of small-payload satellites with characteristically high geometric resolution and higher temporal resolutions (Krieger et al., 2008). This is witnessed by the launch of the Sentinel-1A in 2014, which is part of the Sentinel-1 mission comprising of Sentinel-1A and Sentinel-1B satellites. The Sentinel-1 mission provides a continuous radar imaging and mapping of the Earth with an enhanced revisit frequency, coverage, timeliness and reliability for operational services and applications requir- ing long time series (Malenovsky et al., 2012).

From the above trend, the following data demand- requirements may give an insight into the underpinning

principles: (a) the potential for revisit and situational mon- itoring to cater for short- and long-duration events. For example target based response time for emergency and dis- asters, with the ability for data tasking and acquisition, commitment and delivery; (b) the increasing demand for pre-event and post-event analysis, with the ability of pro- viding answers to the critical scientific questions; (c) the increasing demand for high and very-high spatial and spec- tral resolution images and products in 2D, 3D and 4D geoinformation mapping, visualization and integration in GIS tools in real to near-real time. The EO systems with characteristics and abilities will enable the acquisition of

Y.O. Ouma / Advances in Space Research 57 (2016) 110–126 121

data in different modes such as target data collection, cor- ridor acquisitions, single-pass strip mapping and stereo- and tri-stereo imaging systems as illustrated in Fig. 10.

Fig. 10(a) illustrates the concept of multitarget data col- lection from multisensory platforms for predefined target size and for accurate data acquisition. Similarly, single- pass multistrip mapping in Fig. 10(b)) allows for the strip multiscene and multisensory rapid data acquisition over large areas. A continuous long-band scanning over a large area depicted in Fig. 10(c), is enabled for a predefined cor- ridor. For rapid and more accurate 3-dimensional data acquisition, stereo and tri-stereo focal imaging can be achieved by the observations depicted in Fig. 10(d), from multiple sensors or same sensor but at different orbital locations.

The novel collection modes of the future EO systems will enable high satellite agility that results in a range of responsive modes of operation as depicted in Fig. 11. Such modes of data acquisition can yield super-high spatial res- olutions (SHSR) through across and along track scanning modes as compared to the standard pushbroom modes of data acquisitions (Eves, 2011). With on-demand and high fidelity imagery, coupled with the proposed object-based

(a) Multitarget data collection

(c) Corridor data acquisition

Fig. 10. An illustration on the future on-demand and high-fidelity EO data ac 3D mapping.

image analysis systems, it will be possible to obtain value-added geoinformation for sustainable development.

In addition to the above future EO data acquisition phe- nomena, it is also important to pose the question as to how many EO satellites does the world need in the near future? Given that the Earth is finite, an optimal configuration of EO satellites is possible, covering different ends of spectrum (optical, infrared, thermal, microwave and SAR) and with complementary spatial and temporal resolutions. Given the technological advances in the next 10 years, many coun- tries, some with existing space programs and some with emerging space activities, could contribute to this global concerted effort (Melesse et al., 2007). In recognition of the growing need for improved and streamlined Earth observation activities, governments and international orga- nizations are collaborating through the Group on Earth observations to establish a Global Earth Observation Sys- tem of Systems (GEO-GEOSS) by the year 2015.

A look at the global scenario of EO development, it can be concluded that the 21st century is likely to become the century of interdependence and cooperation between nations in satellite-based imaging of the Earth (Bailey et al., 2001). There is also hope, like in the recent formation

(b) Single-pass multistrip mapping

(d) Stereo and tri-stereo focal imaging

quisition systems with the potential of multiple target imaging for 2D and

Extended strip imaging Highly agile single scene imaging Area mode imaging

Fig. 11. Concept of collection modes for high-fidelity future EO imaging systems.

122 Y.O. Ouma / Advances in Space Research 57 (2016) 110–126

of international collaboration efforts such as the Coperni- cus progamme, and the Committee on Earth Observation (CEOS) constellation for Land Surface Imaging (LSI) (Bailey et al., 2007), all with the aim of building sensing systems and to process the products for end-to-end geoin- formation utility. Similar collaborations and harmoniza- tion are also ongoing in the climate community through the World Meteorological Organization (WMO) in its Glo- bal Space-based Inter-Calibration System (GSICS), WMO Integrated Global Observing Systems (WIGOS) and the Global Climate Observing System (GCOS) programs.

In general, there is the tendency of addressing specific thematic areas of interest such as: hydrology and water, cli- mate change, vegetation monitoring, Earthquake, land- slides and volcanic activities. However at the same time some countries are involved in multithematic areas of EO based resource mapping. A combination of these demands would lead to the requirements for thematic data subsets of a multitude of missions, and none of them would assure a long-term data archiving concept or commit to operational onward distribution (ESA-ESRIN, 2012). As such, collab- oration, harmonization and standardization of EO data for ease of availability and access becoming inevitable.

4.2. Future EO geoinformation data availability and

accessibility

At present, most EO geoinformation data acquisitions are through conventional remote sensing image processing platforms process in stand-alone and centralized platforms. Such systems lack the technique of sharing and distribution information in real-and or even quasi-real-time. The solu- tion with the future forecasts on Big Data in EO is to pro- vide for a centralized and networked mode that is able to allocate and distribute geoinformation products in accor- dance with the desired applications. Because of this chal- lenge, storage, processing, information extraction and distribution or sharing of EO data, rapidly in an effective mode is becoming or has become an issue of concern.

Most geoinformation based applications of EO data require frequent coverage of the same area. This can be maximized by using data from multiple sensors. However, since data from these sensors are acquired in multiple and varied resolutions, they need to be harmonized and synthe- sized before seamless data-sharing. One approach of

coping with this demand is through telegeoprocessing of EO data through standardized, harmonized and central- ized platforms for interoperability (Stoney, 2005).

As defined by Xue et al. (2002), telegeoprocessing deals with real-time spatial databases that are updated on a reg- ular regularly through telecommunications systems. The fundamental goal of telegeoprocessing is to improve on real- or quasi-real time access to geoinformation at afford- able costs for rapid access, processing and decision making. Typically therefore, telegeoprocessing platforms should comprise of WebGIS, mobile geoprocessing and teleGIS. While WebGIS and geoprocessing are familiar concepts, teleGIS refers to the integration of GIS and other telecom- munication techniques. Applications of telegeoprocessing in EO can include environmental telediagnostics and envi- ronmental teleconsultancy (Xue et al., 2002). The adoption of telegeoprocessing will lead to the paradigm shift from centralized systems to wide-area distributed networks for EO data storage, processing and accessibility (Dana et al., 2010).

In addition to telegeoprocessing, grid-computing has also emerged as a computing resource that is based on large-scale resource sharing and high-performance orienta- tion (Foster et al., 2001). For the future EO community, grid-computing will create a platform that can provide access to a wide range of resources for data and high per- formance computing (HPC) software tools. Thus by inte- grating telegeoprocessing and grid-computing, EO analysts and users can dynamically and in a timely manner access the required data for a specific application.

An example of the integrated system for data availabil- ity, computing and access in NASA Earth Exchange (NEX) program, which combines state-of-art super- computing, Earth system modeling, remote sensing data (AVHRR, MODIS and Landsat) as well as global data sets of surface weather records, topography, soils, land cover, global climate simulations and regionally downscaled cli- mate projections. In addition the NEX offers a scientific social network platform to deliver a complete work envi- ronment, which is specific to addressing the climate change and its impacts on the Earth’s global environment. The software components are provided in customizable virtual environments is (Nemani et al., 2011).

Such processed EO data can be accessed from an authoritative data steward through a mapping network as

Constellation of Sensors SCS

Client

Sensor Collection

Service

Map Layers Web Map

Service

Web Feature

Service

Web Coverage

Service

Vector Data

Coverage Data

GetObservation

Measurement Collection

GetCoverage

GetFeature

GetMap

Feature Data

Rendered Map Image

Coverage Data

WCS Client

WFS Client

WMS Client

Coverage Portrayal Service

Styled Layer Descriptor Service

(a)

Web Map Server (WMS)

Web Coverage Server (WCS)

Web Feature Server (WFS)

Data vendor or user

(b) Fig. 12. Schematic framework for advancing data-access and processing through telegeoprocessing and grid-computing platforms with: (a) web-based spatial data processing and collation, and (b) portal for on-demand multiuser and multi-geoinformation data access.

Y.O. Ouma / Advances in Space Research 57 (2016) 110–126 123

schematically represented in Fig. 12. In Fig. 12(a) and (b), different web server clients (WSC) for different geoinforma- tion data access points can be generated and also accessed centrally by data vendors and users. Such WSCs can com- prise of web coverage servers (WCS), web feature servers

(WFS), web map servers (WMS) and EO sensors coverage servers (SCS) as schematically shown in Fig. 12.

Fig. 12 depicts the fact that the enablement of the collec- tion and distribution of accurate and reliable Earth obser- vation information, products and services to consumers

124 Y.O. Ouma / Advances in Space Research 57 (2016) 110–126

worldwide is indeed a complex chain that requires orga- nized architecture. In addition to telegeoprocessing and grid-computing, van Zyl et al. (2009) and Torres- Martinez et al. (2002) argues that sensor webs will in addi- tion allow for the detection of events by EO systems and automatically trigger surveillance by imaging systems.

The demands for shared platforms for data access by data users is already being sought, by users converging on the concept of establishment of shared platforms, sys- tems or facilities where data is made available for access and processing. Conceptualization and investments in such computing platforms are very likely due to the increasing demands, and more so as the economic benefits of EO are being realized. For example the sales in all sectors of application domains and for all the EO data is expected to hit a record high of approximately $4 billion by 2020, with optical data raking in the highest sales at 83%, as com- pared to SAR at 17% of the total sales revenues (Révillon, 2012). This figure has not captured the revenues of from processed data, which contribute to developments and applications in areas related to disaster monitoring and management, global tracking, maritime studies and ocean life, animals and human movement, near-real-time disease epidemic modeling and monitoring, dynamic traffic control and reactive conservation.

5. Conclusions

This paper presents an analysis on the evolution of space-borne Earth observation for land-surface imaging at medium and high resolutions, and seeks to identify the evolution trends, with a view of inferring the new paradigm of Earth Observation in the coming years. This will aid EO users to make an optimum use of future EO systems for the benefit of society. From the review and analysis, it is evi- dent that Earth observation is continuing to grow not only as an operational tool for mapping, monitoring and managing the Earth, and as a primary data source for Earth system science, but also as a socio-economic entity.

Although global issues such as climate change and its effects will continue to provide justification for large multi- sensor satellite projects, the design directions of smaller commercial satellites is rapidly emerging, with focus towards specific application oriented or user defined data acquisition. Analyzing the evolutions of EO systems reveals two paths: one towards small payloads, the other towards large payloads, but both pathways will inevitably lead to higher resolution sensors. With the envisaged advances in EO data acquisition and access platforms, the potentials for real-time and scene-specific imagery is likely to emerge, and there will undoubtedly be personal- ized imagery being beamed, for example, to handheld devices. In the same line of argument, online availability of such imagery will further facilitate real-time and or live visualization platforms.

These developments will stimulate the demand for accu- rate and up-to-date EO data at local- to global-scales. The

resulting large amounts of EO data availability will require new and innovative techniques of dealing with the data intensive issues, which in effect will result into the Earth observation Big Data. This is because as the spatial, tempo- ral and spectral resolutions increases, the data transfer rates and volumes will exponentially grow. This implies that data access, storage and processing will become a dra- matic challenge that users have to cope with through tech- nical equipment and capacity building.

Furthermore, the growing demands for frequent revisits, higher resolution and integration of EO data into GIS, is beginning to call for international coordination for opti- mizing the availability of EO assets. Thus considering future EO systems in the context of Big Data, distributed, web-based processing of geospatial data (telegeoprocess- ing) and grid-computing should be taken into considera- tion. Prerequisites for such systems would call for standardization and harmonization of the EO systems in order to achieve interoperability. Investment in this field is justified by the expanding market and applications of EO information. The implementation of these systems would add-value to the societal and economic benefits of Earth observation, in support of sustainable development.

In summary, the increasing awareness of the benefits of Earth observation in support of sustainable development is continuously stimulating the demand. The increased demand for EO data in turn stimulates and triggers techno- logical development, which leads to improved perfor- mances of EO imaging systems, and further enhances the usability of EO data, in a positive feedback loop on the benefits and the demand. To meet the application and eco- nomic demands, new imaging and sensing systems and modes will emerge that are driven by both operational, application and economic needs.

Acknowledgements

This work was completed when the author was a visiting Alexander von Humboldt Scholar at the University of Applied Sciences in Germany. The author would like to acknowledge the Alexander von Humboldt Foundation for the support. The author is also grateful to Prof. M. Hahn as the host at the Faculty of Geomatics, Computer Science and Mathematics (VIM). Further acknowledge- ments is to the reviewers whose comments immensely improved the paper.

References

Ahlborn, T.M, 2011. Bridge Condition Assessment Using Remote Sensors. Michigan Bridge Conference, March 23 2011. Big Rapids, Michigan, USA.

Avery, T.E., Berlin, G.L.L., 1992. Fundamentals of Remote Sensing and Airphoto Interpretation, fifth ed. Macmillan Publishing Company, New York, pp. 472.

Bailey, G.B., Lauer, D.T., Carneggie, D.M., 2001. International collab- oration: the cornerstone of satellite land remote sensing in the 21st century. Space Policy 17 (3), 161–169.

Y.O. Ouma / Advances in Space Research 57 (2016) 110–126 125

Bailey, G.B., Berger, M., Jeanjean, H., Gallo, K.P., 2007. The CEOS constellation for land surface imaging. sensors, systems, and next- generation satellites XI. In: Roland Meynart, Steven P. Neeck, Haruhisa Shimoda, Shahid Habib (Eds.), Proceedings of SPIE, vol. 6744, 674425, 2007. doi: 10.1117/12.740854.

Borzacchiello, M.T., Craglia, M., 2011. Socio-economic benefits from the use of Earth observation. Report from the International Workshop held at Joint Research Centre, Ispra, 11-13 July 2011. <http://www.crcsi.com.au/assets/Resources/234747a2-1796-45f2-90c3- f57ba435d05d.pdf>.

Che, P.C.L., Zhang, C.-Y., 2014. Data-intensive applications, challenges, techniques and technologies: a survey on big data. Inf. Sci. 275, 314– 347.

Chen, G., Hay, G.J., Carvalho, L.M., Wulder, M.A., 2012. Object-based change detection. Int. J. Remote Sens. 33 (14), 4434–4457.

Coppin, P., Jonckheere, I., Nackaerts, K., Muys, B., Lambin, E., 2004. Digital change detection methods in ecosystem monitoring: a review. Int. J. Remote Sens. 25 (9), 1565–1596.

Dana, P., Silviu, P., Marian, N., Marc, F., Daniela, Z., Radu, C., Adrian, D., 2010. Earth observation data processing in distributed systems. Informatica 34 (4), 463–476.

Del Frate, F., 2007. Use of neural networks for automatic classification from high-resolution images. IEEE Trans. Geosci. Remote Sens. 45 (4), 800–809.

Dos Santos, J.A., Faria, F.A., Calumby, R.T., Torres, R. da S., Lamparelli, R.A.C., 2010. A genetic programming approach for coffee crop recognition. In: IGARSS, pp. 3418–3421, Honolulu, USA, 25–30 July 2010.

Esfandiari, M., Ramapriyan, H., Behnke, J., Sofinowski, E., 2006. Evolution of the Earth observing system (EOS) data and information system (eosdis). In: Geoscience and Remote Sensing Symposium. IEEE IGARSS 2006 International Conference, Denver, Colorado, USA, July 31–4 August. pp. 309–312.

ESA-ESRIN, 2012. Models for Scientific Exploitation of EO data: Innovation, harmonization, technology HMA Workshop Executive Report. 11–12 October 2012, Frascati, Italy.

Euroconsult Report, 2012. Satellite-Based Earth Observation: Market Prospects to 2021. <http://www.euroconsultec.com/report/download? rep=contents&name=satellite-based Earth-observation-2012-toc.42. pdf>.

Eves S., 2011. The future applications of small satellites. Observing the Earth and planets: The Next 50 Years Symposium. 19-21 September 2011, University of Leicester. <http://www2.le.ac.uk/departments/ physics/research/Earth-observation-science/next50years/documents/ next50-session4-eves.pdf>.

Foster, I., Kesselman, C., Tuecke, S., 2001. The anatomy of the grid: enabling scalable virtual organizations. Int. J. Supercomputer Appl. 15 (3), 200–222.

Fritz, S., Scholes, R.J., Obersteiner, M., Bouma, J., Reyers, B., 2008. A conceptual framework for assessing the benefits of a global Earth observation system of systems. IEEE Syst. J. 2 (3), 338–348.

GEO, 2009. Group on Earth observations. GEO 2009-2011 Work Plan Report. http://www.Earthobservations.org/documents/work%20plan/ geo_wp0911_rev2_091210.pdf.

Hay, G.J., Castilla, G., 2008. Geographic object-based image analysis (GEOBIA). In: Blaschke, T., Lang, S., Hay, G.J. (Eds.), Object-Based Image Analysis – Spatial Concepts for Knowledge-Driven Remote Sensing Applications. Springer-Verlag, Berlin, pp. 77–92.

Jacobsen, K., 2005. High resolution imaging satellite systems. e-Proceed- ings EARSeL Workshop 3D Remote Sensing - Use of the third dimension for remote sensing purposes, June 10–11, 2005, Porto, Portugal.

Krieger, G., Gebert, N., Younis, M., Moreira, A., 2008. Advanced synthetic aperture radar based on digital beam forming and waveform diversity. In: IEEE Radar Conference, 26–30 May 2008, Rome, Italy.

Lillesand, T.M., Kiefer, R.W., 2000. Remote Sensing and Image Inter- pretation, fourth ed. John Wiley & Sons, Inc., ISBN: 0-471-25515-7.

Malenovsky, Z., Rott, H., Cihlar, J., Schaepman, M., Garcia-Santos, G., Fernandes, R., Berger, M., 2012. Sentinels for Science: potential of Sentinel-1, -2, and -3 missions for scientific observations of ocean, cryosphere, and land. Remote Sens. Environ. 120, 91–101.

McDonald, R.A., 1995. Opening the cold war sky to the public: declassifying satellite reconnaissance imagery. Photogramm. Eng. Remote Sens. 61 (4), 380–390.

Melesse, A.M., Weng, Q., Thenkabail, P.S., Senay, G.B., 2007. Remote sensing sensors and applications in environmental resources mapping and modelling. Sensors 7 (12), 3209–3241.

Murtaza, H., 2011. Designing a small satellite in LEO for remote sensing application. J. Space Technol. 1 (1), 11–16.

NRO, 2015. US National Reconnaissance Office (NRO). The CORONA Program, CORONA Fact Sheet. http://www.nro.gov/history/csnr/ corona/index.html. Last accessed October 23 2015.

Neeck, S.P., 2015. The NASA Earth science program and small satellites. In: 10th IAA Symposium on Small Satellites for Earth Observation. April 20–24 2015, Berlin Germany.

Nemani, R., Votava, P., Michaelis, A., Melton, F., Milesi, C., 2011. Collaborative Super-computing for global change science. EOS, Trans. Am. Geophys. Union 92 (13), 109–116.

Ouma, Y.O., Ngigi, T.G., Tateishi, R., 2006. On the optimization and selection of wavelet texture for feature extraction from high-resolution satellite imagery with application towards urban-tree delineation. Int. J. Remote Sens. 27 (1), 73–104.

Ouma, Y.O., Tetuko, J., Tateishi, R., 2008. Analysis of co-occurrence and discrete wavelet transform textures for differentiation of forest and non-forest vegetation in very-high-resolution optical-sensor imagery. Int. J. Remote Sens. 29 (12), 3417–3456.

Peri, F., Jr., Hartley, J.B., Duda, J.L., 2001. The future of instrument technology for space-based remote sensing for NASA’s Earth Science Enterprise. Geoscience and Remote Sensing Symposium, 2001. Syd- ney, 09–13 July 2001, pp. 432–435.

Rast, M., Bezy, J.L., Bruzzi, S., 1999. The ESA Medium Resolution Imaging Spectrometer MERIS—a review of the instrument and its mission. Int. J. Remote Sens. 20 (9), 1681–1702.

Révillon, P., 2012. Earth observation market evolution and drivers. In: e- GEOS International Conference, Rome, May 24, 2012.

Sandau, R., Briess, K., D’Errico, M., 2010. Small satellites for global coverage: potential and limits. ISPRS J. Photogramm. 65 (6), 492–504.

Smith, M.W. et al. 2010. ExoplanetSat: detecting transiting exoplanets using a low-cost CubeSat platform. Space Telescopes and Instrumen- tation 2010: Optical, Infrared, and Millimeter Wave. In Jacobus M., Oschmann et al. (Eds.), Proceedings of the International Society for Optical Engineering (SPIE), San Diego, California, USA. v. 7731, SPIE, 2010, pp. 773127–14.

Staenz, K., Held, A. 2012. Summary of current and future terrestrial civilian hyperspectral spaceborne systems, Geo-sciences and Remote sensing Symposium (IGARSS). pp. 123–126. ISSN: 2153–6996.

Stoney, W.A., 2005. Guide to the global explosion of land-imaging satellites; Markets and Opportunities. Earth Imaging Source 2 (1), 10–14.

Stoney, W.E., 1996. The Pecora Legacy – Land Observation Satellites in the Next Century. In the Pecora 13 Symposium, Sioux Falls, South Dakota, 22 August 1996.

Sweeting, M., Hodgson, D., Yi, P., 2012. Advanced Small Satellite Constellations for Earth Observation Services. 33rd Asian Conference on Remote Sensing. Pattaya, Thailand, 26–30 November 2012. 9pp.

Tatem, A.J., Goetz, S.J., Hay, S.I., 2008. Fifty years of Earth observation satellites: views from above have lead to countless advances on the ground in both scientific knowledge and daily life. J. Am. Sci. 96 (5), 390–398.

Torres-Martinez, E., Schoeberl, M., Kalb, M. W., 2002. A Web of Sensors: enabling the Earth Science Vision. In: Proceedings of the International Geoscience and Remote Sensing Symposium, Toronto, Canada, 24–28 June 2002, pp. 351–353.

126 Y.O. Ouma / Advances in Space Research 57 (2016) 110–126

Tzotsos, A., Karantzalos, K., Argialas, D., 2011. Object-based image analysis through nonlinear scale-space filtering. ISPRS J. Pho- togramm. Remote Sens. 66 (1), 2–16.

Van Zyl, T.L., Simonis, I., McFerren, G., 2009. The sensor web: systems of sensor systems. Int. J. Digital Earth 2 (1), 16–30.

Wicks, A., da Silva-Curiel, A., Ward, J., Fouquet, M., 2000. Advancing small satellite Earth observation: operational spacecraft, planned missions and future concepts. In: 14th Annual AIAA/USU Conference on Small Satellites, Logan, UT, Aug. 21–24, 2000.

Wilkinson, G.G., 2005. Results and implications of a study of fifteen years of satellite image classification experiments. IEEE Trans. Geosci. Remote Sens. 43 (3), 433–440.

Wiscombe, W., Herman, J., Valero, F., 2002. L-1 and L-2 Observatories for Earth Science in the Post-2010 Era. In: Proceedings of the International Geoscience and Remote Sensing Symposium, Toronto, Canada, 24–28 June 2002, pp. 365–367.

Xue, Y., Cracknell, A.P., Guo, H.D., 2002. Telegeoprocessing: The integration of remote sensing, Geographic Information System (GIS),

Global Positioning System (GPS) and telecommunication. Int. J. Remote Sens. 23 (9), 1851–1893.

Xue, Y., Li, Y., Guang, J., Zhang, X., Guo, J., 2008. Small satellite remote sensing and applications – history, current and future. Int. J. Remote Sens. 29 (5), 4339–4372.

Zhou, G., 2001. Architecture of Earth observing satellites in Next 10 years and beyond, ISPRS joint workshop on high resolution mapping from space 2001, Sept 19-21, University of Hanover, Germany.

Zhou, G., 2002. An advanced concept on future intelligent Earth observing satellite. In: First International Symposium on Future Intelligent Earth Observing Satellite (FIEOS), November 10–11, 2002, Denver, Colorado, pp. 346–353.

Zhou, G., Jezek, K., 2002. Satellite photograph mosaics of Greenland from the 1960s era. Int. J. Remote Sens. 23 (6), 1125–1142.

Zhou, G., Baysal, O., Kauffmann, P. 2002. Sensor status and future tendency on future Earth observing satellite. In: The First Interna- tional Symposium on Future Intelligent Earth Observing Satellite (FIEOS), Nov. 10–11, 2002, Denver, Colorado, pp. 339–345.