Redefining the possibility of digital Earth and geosciences with spatial cloud computing

Abstract

Global challenges (such as economy and natural hazards) and technology advancements have triggered international leaders and organizations to rethink geosciences and Digital Earth in the new decade. The next generation visions pose grand challenges for infrastructure, especially computing infrastructure. The gradual establishment of cloud computing as a primary infrastructure provides new capabilities to meet the challenges. This paper reviews research conducted using cloud computing to address geoscience and Digital Earth needs within the context of an integrated Earth system. We also introduce the five papers selected through a rigorous review process as exemplar research in using cloud capabilities to address the challenges. The literature and research demonstrate that spatial cloud computing provides unprecedented new capabilities to enable Digital Earth and geosciences in the twenty-first century in several aspects: (1) virtually unlimited computing power for addressing big data storage, sharing, processing, and knowledge discovering challenges, (2) elastic, flexible, and easy-to-use computing infrastructure to facilitate the building of the next generation geospatial cyberinfrastructure, CyberGIS, CloudGIS, and Digital Earth, (3) seamless integration environment that enables mashing up observation, data, models, problems, and citizens, (4) research opportunities triggered by global challenges that may lead to breakthroughs in relevant fields including infrastructure building, GIScience, computer science, and geosciences, and (5) collaboration supported by cloud computing and across science domains, agencies, countries to collectively address global challenges from policy, management, system engineering, acquisition, and operation aspects.

Keywords:

• EarthCube
• CloudGIS
• eScience
• Earth science
• interoperability

1. The challenges of geosciences and Digital Earth

Technology advancements and globalization require the integration of previously separate fields, including geography, geochemistry, hydrology, environmental sciences, climate science, soil science, and others in the twenty-first century (National Research Council [NRC] 2012a) to (1) address international challenges, (2) obtain fresh approaches for solving complex problems, and (3) protect local, regional, national, and international interests in the context of an integrated Earth system. The integrated system also contributes significantly to the human capability in answering many previous unanswerable scientific enquires, such as those from climate and land use change, ecosystems, energy and minerals, environmental health, water, and natural hazards. These existing and emerging international science opportunities as characterized by US Geological Survey include (NRC 2012a) (u1) global natural hazards planning and responses, (u2) global energy and mineral resource assessments, (u3) enhanced water sustainability research in desert regions and tropical areas, (u4) use of climate and land-cover science for decisions on climate adaptation and natural resource management, (u5) understanding the influence of climate change on ecosystems, populations, and disease emergence, (u6) clarification and development of invasive species work using trade patterns, refugee situations, and changing climate and environment for initial prioritization, (u7) quantitative health-based human health risk assessment analysis based on contaminant exposure levels, (u8) ecological and quantitative human health risk assessment analysis based on contaminant exposure levels, (u9) research in water contamination and supply, (10) water and ecological science in cold regions sensitive to climate change, and (u11) comprehensive enhancement of, and accessibility to, essential topographic and geologic map information through the topological and geological modeling and mapping.

To conduct more in-depth research, National Science Foundation (NSF) also worked with NRC (2012b) to identify the new research opportunities in geosciences for the next decade as (n1) early Earth, (n2) thermo-chemical internal dynamics and volatile distribution, (n3) faulting and deformation processes, (n4) interactions among climate, surface processes, tectonics, and deep Earth processes, (n5) co-evolution of life, environment, and climate, (n6) coupled hydrogeomorphic-ecosystem response to natural and anthropogenic change, (n7) biogeochemical and water cycles in terrestrial environments and impacts of global change, and (n8) recent advances in geochronology.

Among the new research frontiers identified for USGS and NSF, u1 and u11 are unique to USGS and n1, n2, n3, and n8 are unique to NSF. The following are complementing pairs: u2-u3 with n4, u4-u6 with n5, u7-u8 with n6, and u9-10 with n7. In this regard, although the two sets of geoscience challenges have different focuses, they complement each other and both identify a priority for training the next generation geoscientists. And this complementary is considered in Table 1.

Table 1. Geoscience vision and the spatial cloud computing research (u1–u11 and n1–n8 are in section 1 and adopted from NRC 2012a, 2012b).

	USGS	u1	u2	u3	u4	u5	u6	u7	u8	u9	u10	u11
Technologies\vision	NSF		n4		n5			n6		n7			n1	n2	n3	n8
Architecture		X			X	X		X	X			X
Visualization		X	X	X	X	X	X	X	X	X	X	x	X		X
Big data		X		X	X	X	X	X	X			X		X
Real-time		X		X				X	X	X				X
Data storage		X	X	X	X	X	X	X	X		X	X				X
Co-location		X		X	X		X	X	X	X	X	X
Spatial methods		X	X	X	X	X	X	X	X	X	X	X	X	X	X
Interoperability		X		X	X	X	X	X	X	X	X	X		X
Event		X		X	X	X	X	X	X	X	X		X	X	X
Vision					X			X	X
Reliability		X		X	X			X	X			X		X
Cloud management		X		X	X			X	X			X
Communication		X	X	X	X	X		X	X		X
Security		X	X		X	X	X	X	X	X
Spatiotemporal		X		X	X	X	X	X	X	X	X		X	X	X	X
Collaboration		X	X	X	X	X	X	X	x	X	X	X		X

While these national and international deep scientific inquiries being addressed, international efforts are put on the advancements of the technologies to support the scientific discoveries and applications driven by the Digital Earth concept (Gore 1998) and recently, by the NASA Earth Exchange (NASA 2012) and NSF (2011) EarthCube effort at a US national level and the intergovernmental Group on Earth Observation (GEO 2012) at the global level. Under the auspices of the International Society of Digital Earth, relevant activities defined the research directions and challenges towards 2020 as a next generation Digital Earth by considering the recent development and challenges (Craglia et al. 2012; Goodchild et al. 2012). Craglia et al. (2012) defined five themes for next generation Digital Earth as (d1) a research challenge, (d2) an information system, (d3) applications, (d4) organizational metaphor, and (d5) a strategic infrastructure.

All these visions call for the readiness of digital technologies for a computing infrastructure to support the scientific quests and application needs (such as Gore 1998; NSF 2011; Craglia et al. 2012; NASA 2012). The development of sharing and utilizing computing resources across geographic boundaries provided ideal methodologies to solve the problem and construct such an infrastructure (Yang et al. 2010). The sharing of distributed computing has evolved from early High Performance Computing (HPC), grid computing, peer-to-peer computing, and cyberinfrastructure to the recent cloud computing, which realizes access to distributed computing for end users as a utility or ubiquitous service (Yang et al. 2010). Research for adopting cloud computing to enable or solve the geoscience problems and Digital Earth challenges have also attracted many computational scientists to investigate the readiness of cloud computing (as reviewed in section 2 and section 3). Researches were also conducted to explore how the spatiotemporal principles that govern the geosciences and Digital Earth can be utilized to optimize the cloud computing and provide better computing solutions via spatial cloud computing (Yang 2011). CloudGIS was also brought up as the next generation GIS to provide GIS software and functionalities through cloud computing platforms (Wu and Wu 2011).

After the publication of the spatial cloud computing definition paper (Yang et al. 2011), we edit this spatial cloud computing special issue to capture the latest investigations of global scientific communities in utilizing cloud computing to address the geoscience and Digital Earth challenges and to identify the future research directions in this emerging field. The next section summarizes, in examples, research conducted to utilize cloud computing for enabling geoscience and Digital Earth. Section 3 reviewed the technological advancements for addressing the scientific and application problems. Section 4 introduces the selected papers, section 5 concludes the paper with a recommended list of research agenda for spatial cloud computing to support a new generation of geospatial information system (CloudGIS), and section 6 analyzes how the cloud computing research and capabilities can be utilized to redefine the possibility of Digital Earth and geosciences.

2. Geosciences enabled by cloud computing

The geosciences enabled by cloud computing include almost all dimensions of the new geosciences as defined by USGS (NRC 2012a) and NSF (NRC 2012b), for example:

1.	Early Earth is one of the most challenging subdomains. Research has been conducted to enable scientists access easily to astrophysics simulation and visualization through science gateway supported by cloud computing (Pajorová and Hluchý 2011a) and enables the processing of astronomy data to search for Earth-like planets orbiting other stars by integrating several computational clouds including FutureGrid, NERSC's Magellan cloud, and Amazon EC2 (Vöckler et al. 2011). Early Earth research is generally centralized but the computing is distributed in a SETI@Home fashion with a relative simpler demand for computing infrastructure.
2.	Energy and mineral science requires a good data management to support modeling of the generation and distribution of energy. Liu, Wang, and Liu (2012) used cloud computing to address the data, storage, and processing demands for energy information management. Energy and mineral science is relatively complex with very broad spatiotemporal distribution of producers, managers, and consumers of material/information, and demands a relative complex computing infrastructure.
3.	Climate science faces the grand challenges of big data management and analysis. Cloud computing has been used to enable the effective management of large-scale collections of observational data and model output data for community-defined services such as the Earth System Grid (Schnase et al. 2011). Huang, Gangl, and Bingham (2011) used cloud computing for climatology services of storing and analyzing spatiotemporal characteristics of scatterometer data over Antarctica. Tran et al (2011) used cloud computing and Apache Object Oriented Data Technology synergistically to form an effective, efficient, and extensible combination to the challenges of NASA science missions' data management at reduced costs. The integration of climate science and environment, ecological, and health sciences will make it increasing complex in the next decades.
4.	Traffic management and simulation systems require the (near) real-time capabilities enabled by cloud computing to (1) solve data-intensive geospatial problems in urban traffic systems for traffic surveillance management (Li, Zhang, and Yu 2011), (2) integrate the management of International Traffic Database and support project communication and publishing (Miska and Kuwahara 2010), and (3) provide the most environmentally friendly transport solution with intuitive and individualized services for a dynamic number of end users (Di Martino, Giorio, and Galioro 2011).
5.	Ecology faces the challenges of storage, scalability, and platform integration in a global context. Cloud computing and open source has been utilized to (1) enable storage, scalability, and deployment flexibility for global marine biogeography data and analyses (Fujioka et al. 2012), (2) setup a platform for forest pest control to handle the huge amount of pest data (Jiang et al. 2010), (3) provide worldwide integrated monitoring of the environment and its inhabitants, understand their interrelationships, improve our ability to protect the planet and its people by integrating hundreds of thousands of data sources (Montgomery and Mundt 2010), and (4) support sustainability research (Mobilia et al. 2009).
6.	Civil engineering and water management is critical for human being, and Behzad et al. (2011) used both HPC and cloud computing in a hybrid cyberinfrastructure to support groundwater ensemble runs for forecasting the availability of fresh water.
7.	Disaster/waste management, disaster monitoring, forecasting, warning, preparation, and response can be supported efficiently by cloud computing (Liang, Lii, and Chang 2011). For example, Bessis, Asimakopoulou, and Xhafa (2011) use the next generation emerging technologies for enabling collective computational intelligence in managing disaster situations. Ishikawa, Sugiyama, and Sasaki (2011) investigated using cloud computing and satellite images to monitor and simulate the dispersion of industrial waste to reduce waste impact.
8.	Human and environment health is another example needing global computational flexibility and extensibility that can be provided by cloud computing for prediction analyses (Bohm, Mehler-Bicher, and Fenchel 2011). Shen et al. (2012) used service-oriented architecture and cloud computing to support analysis and visualization of medical data highlighting the global variation of health data by geography, living habits, and cultures. Eriksson et al. (2011) used a hybrid cloud-based simulation architecture for pandemic influenza simulation and found that it is possible to develop a scalable simulation environment using cloud computing.

Among these examples, the most popular domains (such as energy, transportation, civil engineering, disaster/waste management, and human and environment health) will benefit the most from cloud computing to support the massively distributed and concurrent end-user requests with elasticity, on-demand, and pay-as-you-go features (Yang et al. 2011).

3. Key technology advancements for cloud computing

Besides the virtualization, web services, and service-oriented architecture technologies, the driving fundamental technologies of cloud computing include (1) parallel computing, (2) multitasking supported by multiple computers and cores, and (3) distributed or collaborative processing through the natural distribution of data, problems, computations, and users. There are many efforts to explore the parallelization aspect of cloud computing. For example, Tilevich and Eugster (2010) organized a workshop to explore programming support innovations to address the emerging distributed applications and the state-of-the-art of their programming support. Akdogan et al. (2010) studied the problem of parallel geospatial query processing using the MapReduce programming model with spatial index and Voronoi diagrams. Wang and Liu (2008) researched parallel computing architecture structure based on cloud computing for parallelizing data-mining algorithms. Zhang (2010) developed a parallel spatial statistics module for visual explorations on top of Personal HPC-G in combination with cloud computing. van Zyl et al. (2012) extended multitasking and distributed processing capabilities for Earth observation scientific workflows in a distributed computing environment to allow these geospatial processes to be seamlessly executed across distributed resources. Karimi, Roongpiboonsopit, and Wang (2011) studied distributed algorithms for geospatial data processing on clouds and compared their experimentation with an existing cloud platform to evaluate its performance for real-time geoprocessing. Panchul, Akopian, and Jamshidi (2011) reported that depending on the applications, the best possible results are produced by different parallelization approaches from hardware-implemented parallelism to software multithreading. Considering the driving principle and the requirements of geosciences and Digital Earth, the enablement by cloud computing lies in the advancements of several key techniques:

1.	System architecture is the key to a successful computing platform. Cloud computing is the latest success of the distributed computing architectural paradigm after HPC and Grid Computing (Mateescu, Gentzsch, and Ribbens 2011). Current solutions utilizing cloud computing may benefit from the integration of a hybrid framework of HPC, grid, cloud, and cluster computing to solve problems of large-scale sciences such as Earth science, astronomy, and related sciences (Pajorová and Hluchý 2011b). An optimization strategy is also proposed (Cui et al. 2011) to unify the management of spatial data from multiple sources using cloud computing and existing legacy systems. Rothenberg (2010) reviewed developments that have taken place in architecting data center networks to meet the requirements of the cloud and speculated on the potential impacts of such computing developments in shaping the future Internet by driving incentives of adoption of new protocols and architectural changes.
2.	Visualization is a key to the success of cloud computing by providing easy-to-use interfaces to domain specialists without advanced technical knowledge (i.e. meteorologists, geography specialists, hydrologists, etc.) to manipulate Big Data in Earth observation and geosciences (Stefanut, Popescu, and Gorgan 2011). Many research efforts, such as that by Pitcher (2009), have been conducted to investigate cloud computing in combination with virtual Earth and computer graphics to provide easy-to-use visualization interfaces for end users.
3.	Big Data is a popular challenge that cuts across all geoscience and Digital Earth subdomains. The advancement of managing and processing Big Data can be greatly enhanced by using cloud computing (Karimi, Roongpiboonsopit, and Wang, 2011) to support geoprocessing for real-time navigation applications. On the other hand, current cloud computing platforms require improvements and special tools for handling efficiently real-time geoprocessing, such as iGNSS QoS prediction (Karimi, Roongpiboonsopit, and Wang, 2011).
4.	Real-time data processing is required by many geoscience and Digital Earth applications. Cloud computing could enable the real-time response with virtually unlimited resources and on demand services within minutes for applications, such as geo-streaming (Kazemitabar, Banaei-Kashani, and McLeod 2011) and the GEOSS clearinghouse (Huang et al. 2010).
5.	Data storage is another challenge related to Big Data support. Cloud computing can provide inexpensive and simple solution for users to post and share data on the web (Bunzel, Ager, and Schrader-Patton 2010). Jiìíček and Di Massimo (2011) reported users can easily upload and store public data into the Microsoft Cloud, while leveraging the Windows Azure Platform and environment for processing. However, data storage and management need additional research to reduce the hosting cost for inherently large volumes of Earth Science data and maintaining an easy-to-use experience.
6.	Data and process co-location is a key optimization approach that may improve cloud throughput. This means the data processing can be conducted at different places according to scheduling strategy by either shipping the data or the processing modules to address the significant latency arising from frequent access to large datasets and corresponding data movement between distributed data centers (Deng et al. 2011). Liu et al. (2011) argued that the most related datasets can be placed into the same data center based on the data dependence at workflow build-time; the tasks are then scheduled to their most closely related data centers for execution and the newly generated data-sets are put into the data center that has the most dependency at workflow runtime.
7.	Key spatial methods must be implemented in the cloud to provide basic support to Digital Earth and geosciences (Goodchild et al. 2012). Key methods, including spatial data storage, spatial indexing, and spatial operations, should be researched systematically with respect to optimal use of cloud computing (Wang and Wang 2010). The combination of cloud computing with geoscience and Digital Earth problems could also generate new key spatial methods, such as optimal spanning tree for securing distributed data replicas in geographically dispersed clouds (He et al. 2012) and spatiotemporal indexing to improve cloud application efficiency in a global or regional scale.
8.	Standardization and interoperability are keys to integrate systems across diverse cloud platforms. For the geoscience and Digital Earth applications, the standards and interoperability sit in the Web service interface level of cloud computing architecture. Examples include investigation into the best cloud deployment approach to support the Open Geospatial Consortium (OGC) Web Coverage Service specification (Shao et al. 2011), how OGC Web service standards are becoming ubiquitous in cloud computing environments replacing old computing models for geoprocessing (Reichardt 2010), and how the application interoperability bottleneck in a data-intensive application can be solved by cloud computing (Wu, Wu, and Huang 2010).
9.	Event detection and computing is essential for best utilizing cloud computing resource to support autoscaling for elasticity and utility characteristics. Detecting events in cloud computing platforms (Helmer, Poulovassilis, and Xhafa 2011) will require new spatiotemporal algorithms to exploit the scalability of cloud computers while working within the limits of state synchronization across multiple servers in the cloud (Olson et al. 2011).

These technological advancements will pave routes for implementing cloud computing and enabling geosciences and Digital Earth. The full implementation of the five National Institute of Standards and Technology (NIST 2011) cloud computing characteristics will further the advancement and foundation building of the cloud computing for geosciences and Digital Earth.

4. Introduction to the papers

We called for and received 25 abstract submissions and invited 10 full paper submissions. Five papers were selected for publication based on a rigorous peer review process.

Liu et al. (2013) reported research using Microsoft Azure to support the on-demand requirement of groundwater supply analyses using Texas and Arizona examples. Technology and license problems were solved by integrating Azure and Dropbox for computing, integrating desktop and cloud for software license access, and transferring files between the desktop and the cloud for data sharing. They found that cloud computing could enhance groundwater analyses by providing informed uncertainty analysis results that assist groundwater planning and sustainability analysis.

Dust storms are a challenging natural hazard facing us in the twenty-first century with an increasing frequency and broad geographic and socioeconomic impacts. The prediction of dust storms involves modeling uncertainty and requires significant computing resources. Taking the computing challenge and exploring spatiotemporal optimization (Yang et al. 2011), Huang et al. (2013) utilized cloud computing to provision large quantities of computing resources on-demand for a short time period to support high-resolution dust storm forecasting over large geographic region, while reducing the cost of forecasting by exploiting the measured service and elastic capabilities of the cloud environment.

Wen et al. (2013) reported using cloud computing to support an open environment for sharing geographic analyses models. To solve the heterogeneity problem, they used several strategies of model description, model encapsulation, model deployment, and transparent access, and verified the strategies with an experimental environment established on a cloud computing platform. They also identified several future research challenges including interoperability, performance, system security, load balancing, and quality of service assessment.

Following the parallelization spirit, Kim and Tsou (2013) compares cloud computing approaches to those of grid computing using WebGIS-based geoprocessing simulations. They found that with limited amount of parallelization, grid computing has a better performance but with increased parallelization, cloud computing performance is comparable. Their research proves cloud computing to be a viable solution for enabling computation and data-intensive processing for complex GIS models. From an accessibility aspect, the on-demand aspects of commercial cloud computing allows timely access at low cost. They also commented that different instances of cloud computing can satisfy different WebGIS computing needs; therefore, assessments should be conducted to evaluate the best-fit cloud instance for specific applications.

Yue et al. (2013) provide a comparative analysis of the design and implementation of geoprocessing services in Microsoft Azure and Google App Engine. The research compares the running environment, programming language, application framework, storage service, and platform application programming interfaces for both cloud platforms for geoprocessing functions. The result provides the reference on selectively utilizing cloud computing platforms in a hybrid cloud pattern. The research shows that virtualization is the key for portable geoprocessing cloud services. The performance tests demonstrate how the cloud computing can support the on-demand geoprocessing and economic deployment of geoprocessing services.

In addition to the five accepted papers, we added this field review paper, fully reviewed by a broader expert base and the community, to capture the state-of-the-art and to identify the future research directions.

5. Research directions and towards a research agenda

Cloud computing provides enabling capabilities for geosciences and Digital Earth in the twenty-first century. The eventual success of the spatial cloud computing and the next generation GIS – CloudGIS – will be determined in five aspects (Luftman and Zadeh 2011): (1) improvements to research productivity and cost reduction; (2) alignment to current information technology (IT) procedure; (3) agility and speed in responding to computing needs; (4) ease of application in scientific research; and (5) improvement in the reliability and efficiency of IT. As global collaboration requires, cloud computing in support of geosciences and Digital Earth will transcend organizations, jurisdictional boundaries, and continents (NRC 2011b). The potential impediments and solutions can be addressed relative to policy, management, system engineering, acquisition, and operations requirements to ensure the eventual success of CloudGIS. The following cloud computing research directions need additional attention from the interdisciplinary domains:

1.	Continuing visions for geosciences and Digital Earth will provide driving demands for cloud computing advancements. These visions could include, to name a few, smart/intelligent Earth (Liu et al. 2010) and campus (Liu, Xie, and Peng 2009); virtual geographic environment (Lin et al. 2011); global military conflict simulator (Tanase and Urzica 2009); global sharing of Earth observation data (Huang et al. 2010); new cartography (Meng 2011); and security, data, and future of computing (IEEE 2011).
2.	Cloud interoperability has to rely on the standards developed by different organizations, such as OGC, OGF, NIST, ISO, and IEEE, through a systematic architecture designed to support the sharing of distributed computing resources at different levels. This has to be driven by large user groups, application domains, vendors, and governments to achieve the required level of interoperability (Lee 2010).
3.	New visualization and interactive systems for cloud computing will be essential to support ease-of-use and straightforward access (Vogel 2011). For example, regional arctic systems modeling (Roberts et al. 2010) and a polar research cyberinfrastructure (Yang, Nebert, and Taylor, 2011) require a computing infrastructure that is easily accessible, extensible, and usable.
4.	Reliability and availability is a challenge to achieve in a cloud platform where most components are distributed and independently managed. This includes being able to access the platform from different regions and the efficiency of the access.
5.	Real-time simulation and access is essential for different types of decision support from emergency response to individual decisions. Relevant research should include theory-based simulation and multiscale, multicomponent modeling, as well as data-intensive and interactive visualization capability for both cloud computing platforms and applications (NRC 2011a).
6.	Cloud management needs significant improvement to meet the five characteristics defined by NIST (2011). Systematic management at the enterprise level to support semiautomatic engineering operations for the cloud and the entire information system will be essential (Choi and Lee 2010).
7.	Cloud outreach should help convey the message to (potential) users properly with its own strengths and shortcomings.
8.	Security is a big concern when utilizing cloud computing to deploy a distributed platform with certain information, for example, for labor and social security applications (Lu 2010). Examples include a secure web environment to develop phenologic matrices from linked media in real time (O'Leary and Kaufman 2011) and creation of international license agreements or exceptions to ensure that export-controlled technical data stored on the cloud is secure and protected (Schoorl 2012). It will be a continuing challenge in how to ensure the protection of sensitive data, privacy, and systems while maintaining the sharing spirit of cloud computing.
9.	Spatiotemporal optimization is key to fully realize the benefit of cloud computing to geoscience, computing infrastructure, Digital Earth, and education (Yang et al. 2011). Addressing fundamental science questions and application problems and optimizing the cloud computing platforms with spatiotemporal principles will help lay the foundation to implement spatial cloud computing (Yang et al. 2010).
10.	Global collaboration is important to make a difference on several fronts: (1) the integration of multiple scientific domains through common information science platforms will require global collaboration (Yang et al. 2010); (2) GIScience advancement will depend on global collaboration from the identified directions of (g1) position technology for knowing where everything is, at all times, (g2) citizen science and crowd-sourcing, (g3) dynamic or spatiotemporal events, (g4) the 3rd, 4th, and 5th dimensions, and (g5) the challenge of education (Goodchild 2010); (3) Craglia et al (2012) also emphasized the importance of global collaboration for the next generation Digital Earth in the light of the many developments from IT, data infrastructures, and Earth observation. It is essential to develop a series of collaborations at the global level to turn the vision into reality.

In addition to this list, other research frontiers should be added when needed. For example, cloud-based knowledge discovery will require data mining and knowledge discovering across linked data distributed worldwide and without fully accessing the actual data.

6. Enabling the visions with cloud technologies

This section provides an analysis of how the spatial cloud computing advancements and future research would enable the visions. Two levels are identified as kernel support (marked with x) or lightly dependent (without mark). This may serve as a reference in conducting research of spatial cloud computing to enable the visions. The technologies are detailed in section 3 and section 5.

Table 1 illustrates that natural hazards, water sustainability, climate adaptation and natural resource management, environmental and human health will be very comprehensive and require advancements in all different technological aspects to address the problems. It also illustrates that visualization, spatial methods, and spatiotemporal studies will be key areas for most geoscience domains.

Table 2 illustrates that Digital Earth as an infrastructure will foster the advancements of different technological aspects of cloud computing. Big Data, event, and cloud management will be the key technologies for enabling Digital Earth and geosciences.

Table 2. Digital Earth vision and the spatial cloud computing research (d1–d5 are in section 1 adopted from Craglia et al. 2012).

Technologies\vision	d1 research	d2 system	d3 application	d4 metaphor	d5 infrastructure
Architecture	X	X	X		X
Visualization			X	X	X
Big data	X	X	X	X	X
Real-time	X	X	X		X
Data storage		X	X		X
Co-location	X	X	X		X
gis methods	X
Interoperability		X		X	X
Event	X	X	X	X	X
Vision			X	X
Reliability		X	X		X
Cloud management	X	X	X	X	X
Communication				X	X
Security	X	X			X
Spatiotemporal	X	X	X	X
Collaboration	X		X	X	X

Table 3 illustrates that education will be critical for cloud computing. Geovisualization, events, and spatiotemporal studies will be the key areas of GIS that enables the advancements of cloud computing.

Table 3. GIScience vision and the spatial cloud computing research (g1–g5 are in 10th item of section 5 adopted from Goodchild 2010).

Technologies\vision	g1. position	g2. citizen science	g3. dynamics	g4. 3+ dimension	g5. education
Architecture			X	X	X
Geovisualization	X	X	X	X	X
Big data		X	X	X	X
Real-time	X		X	X	X
Data storage		X	X	X	X
Co-location	X		X	X	X
gis methods	X		X	X	X
3Interoperability	X		X	X	X
Event research	X	X	X	X	X
Vision		X	X	X	X
Reliability	X			X	X
Cloud management		X	X		X
Communication		X			X
Security	X	X			X
Spatiotemporal	X	X	X	X	X
Collaboration	X	X			X

Acknowledgements

Research is supported by State Administration of Foreign Experts Affairs (20120464001), NSF (IIP-1160979 and CNS-1117300), FGDC (GeoCloud and GEOSS Clearinghouse), and Microsoft Research. Drs. Dennis Guo, Xiang Li, Ziyong Zhou, Yang Hong, Peng Yue, Rick Kim, Yong Liu, Qunying Huang, Min Chen, Xinyue Ye, and Santonu Goswami reviewed the manuscript. We sincerely thank Drs. Huadong Guo and Changlin Wang for inviting us to organize the special issue and facilitating the process of developing this special issue.