Populating digital earth: improving access to Chinese remote sensing data for terrestrial applications

Abstract

Global change has a significant impact on the lives of humankind. Earth observation can help to better understand our earth and cope with global change. With the availability of more reliable environmental data sets, digital earth is becoming a popular way to monitor the Earth and provide information to researchers and decision makers on environment protection, disaster mitigation, and social benefits. Therefore, accessing data with lowering costs is essential for digital earth. Nevertheless, there are big challenges in ensuring the feasibility of access to Chinese remote sensing data. This paper outlines some of the main challenges in realizing data sharing, provides an analysis of the core reasons leading to these challenges, and proposes recommendations to overcome the challenges. Amongst the main challenges are differences in data policy to gain access to satellite data, diverse data formats, and delivery mechanisms. The major challenge for the decision makers is to define a more open policy and for the scientist the challenge is to implement these polices for the benefit of all. This paper proposes that governments should adopt policies encouraging more open distribution and access to their data, in order to generate an improved digital earth with increased benefits to human society.

Keywords:

• date policy
• remote sensing
• data sharing
• China

Importance of data sharing policy for government remote sensing data

The vast majority of environmental remote sensing satellites have been launched by governments, though the commercial sector has become more important in the last decade. The main purposes of the resultant data for government are to meet scientific needs relating to global change, to provide support for international environmental agreements, and to underpin decision making relating to management of natural resources and disasters.

Remote sensing technology has become a crucial tool for earth observation and global change monitoring. The ability to easily access remote sensing data and have user-friendly data systems is a significant requirement for every country. Therefore, developing proper policies for government remote sensing data has enormous impacts not only on the efficacy of observation systems locally and globally but also on their social and economic applications. Given the importance of data policies in every country, governments have either developed country-wide policies but often, as in China, individual government agencies have formulated their own data policy based on their individual needs and also on their underlying beliefs about whether to charge users.

Government agencies and departments have the responsibility to provide information and data to meet public needs and scientific interests. This relates especially to the needs to understand environmental and climate change; also government agencies and departments have responsibility to provide information and data to decision-makers to help them formulate and implement policies at national, regional, and local levels. On the other hand, national security issues may be of concern, countries may wish to restrict open access to data due to concerns of national security. This conflict between better understanding of the earth to cope with issues such as global climate change and restricting data access for national security concerns is present in most countries with remote sensing capabilities, but there is increasing openness in the sharing of data.

We argue in this paper that free and open sharing of data should be supported. From historical experiences we show openness of data sharing leading to better observing systems and improved understanding of the environment at all scales. Open sharing of data to all for free or at minimal cost will increases use, builds applications, and value-added industries; in contrast, charging and restricting access inevitably reduces use and hinders applications and the growth of value-added industries. We discuss these issues in the context of contemporary Chinese remote sensing capabilities and current data sharing policies.

Chinese terrestrial remote sensing satellites

China has been developing an extensive set of satellites for various types of environmental monitoring for over 20 years. At present there are four series of national satellites operated by China, which include earth resource satellites, environmental monitoring satellites, meteorological satellites, and marine satellites. Apart from these national satellite programs, China also has some small satellites owned by private companies and universities, such as Beijing-1 satellite and ‘Hangtian Tsinghua-1’ Micro-Satellite. China is developing a comprehensive set of remote sensing systems from data collection by earth observation satellites, through data reception processing and delivery systems to enable useful applications to be informed. A comprehensive description of Chinese satellite capabilities is provided by Guo (2012). The account below provides an updated summary of this review.

Earth resources satellites

China is developing Earth resource satellite in cooperation with Brazil. The first two satellites, named as CBERS-1 and CBERS-2 under the Chinese Brazilian Earth Resources Satellite (CBERS) series were launched in October 1999 and October 2003, respectively. Both satellites were equipped with 20-m resolution charge-coupled device (CCD) cameras, an Infrared Multi-Spectral Scanner (IRMSS) with resolution from78 to 156 m, and a Wide Field Imager (WFI) with 258 m resolution. CBERS-2B, launched in September 2007, was equipped with a 2.36-m resolution High Resolution Camera in addition to CCD and WFI sensors similar to those on CBERS-1/2. Based on the experience of CBERS-1, CBERS-2, and CBERS-2B, the resource satellite named ZY-1-02C was launched in December 2011. ZY1-02C is equipped with high resolution camera and sensors covering the full spectral range from visible to microwave wavelengths. Resource III satellite named ZY-3 was successfully launched on 9 January 2012 (CRSDA 2012). Resource III satellite weighs about 2650 kg, with a design life of about five years. The main task of the satellite is to achieve a national coverage of long-term, continuous, stable, rapid acquisition of high-resolution three-dimensional images and multispectral images. It is equipped with CCD forward and backward cameras with 6m resolution (CRSDA 2012). Subsequent satellites will be improved by adding a multi-spectral (10–20m) CCD camera with a 5-m panchromatic band. The swath width will be 120 km and 60 km, respectively; there will also be an IRMSS and WFI (866 km swath width) payloads.

Environmental monitoring satellites

Environmental monitoring satellites have similar characteristics to the earth resource satellites but with higher revisit capabilities to make them more suitable for environmental and disaster monitoring. China launched Shijian-5 in 1999, which is based on a multi-purpose, 3-axis stabilized small satellite platform developed by China. Based on the platform, China is launching a small satellite constellation for environment and disaster monitoring known as the HJ series of satellites (sometimes also called environment and disaster monitoring satellites). The HJ series consist of four optical satellites and four synthetic aperture radar (SAR) satellites; the orbit of the constellation is sun synchronous, with a revisit time of 96 hours with one satellite or 48 hours with two satellites. The payloads of the satellites include a CCD camera, infrared camera, hyper-spectral camera, and an S-band SAR (Table 1). The optical sensors have a 650 km swath with a 10:45AM sun-synchronous orbit (Wang et al. 2005).

Table 1. HJ satellites and the technical specifications (Guo 2012).

HJ satellites	Payload	Orbit	Spatial resolution (m)	Swath width (km)	Revisit cycle (days)
HJ-1A	CCD Hyperspectral Imager	sun synchronous	30 100	360 (1 camera) 700 (2 cameras) 50	4
HJ-1B	CCD Infrared Multispectral Camera	sun synchronous	30 150–300	700 720	4 4
HJ-1C	Synthetic Aperture - Radar	sun synchronous	5 (single look, strip mode)/20 (4 looks, scan mode)	40-stripe mode/ 100-scan mode	4

The four optical satellites and four SAR satellites were originally scheduled to be launched between 2008 and 2012. The first three satellites consist of two optical satellites and a SAR satellite, called the ‘2 + 1’ Project. The first two satellites HJ-1A and HJ-1B were launched in September 2008. HJ-1C was launched in November 2012. Further optical satellites are expected to be launched soon. But the launch dates are still unclear.

Meteorological satellites

Although primarily designed for meteorological applications in practice, such satellites are often used for terrestrial applications. Chinese meteorological satellites are named as FY (Fengyun – literally wind cloud). FY followed by even numbers indicate geostationary meteorological satellites like FY-2 and FY-4 series; whereas FY followed by odd numbers are polar-orbiting meteorological satellites, such as FY-1 and FY-3 series (Zhang et al. 2006).

Following the FY-1 series (launched on 7 September 1988), satellites in the FY-3 series are the second generation Chinese polar orbiting meteorological satellites (Zhang et al. 2006). FY-3A was launched in May 2008. FY-3B was launched in November 2010, which carries nine payloads to provide global, all-weather, multi-spectral images and sounding capabilities to provide quantitative observations and services. The FY-3 series will be advanced meteorological satellites with a spatial resolution of up to 250m. The World Meteorology Organization (WMO) has agreed to accept the series as part of the WMO operational meteorological satellite constellation for 2005–2020.

The FY-4 series will be the second generation of Chinese geostationary meteorological satellites after the FY-2 series, which is planned to be launched after 2010 (Zhang et al. 2005). It will use a 3-axis stabilized satellite platform to increase its capability for more frequent observations of selected areas. It will also have more imaging scanner channels, an infrared vertical atmospheric sounder, and a lightening imager (Table 2).

Table 2. FY satellites and the technical specification (Guo 2012).

Meteorological satellites	Spectral ranges	Spatial resolution (m)	Swath width (km)	Launch time
FY-1A/B	VIS/NIR/TIR	1100/4000	2860	06.09.1988/03.09.1990
FY-1C/D	VIS/IR –	1100/4000 –	3100 –	10.05.1999/15.05.2002
FY-2A/B/C/D/E	VIS/IR	1250/5000/5760	–	10.06.1997/25.06.2000/ 19.10.2004/08.12.2006/ 23.12.2008
FY-3A/B	VIS/IR EHF/U-band X/Ku/K/Ka/W-band UV/VIS/IRUV	17 km/1100/250-1000 15 km/50–75 km 15–85 km – 200 km/50 km	2800 2700 1400 –	27.05.2008/04.11.2010

Note: EHF, extremely high frequency; IR, infrared; NIR, near infrared; UV, ultraviolet; VIS, visible.

Small satellites

The ‘Hangtian Tsinghua-1’ micro satellite was launched in June 2000 and Beijing-1 satellite was launched in October 2005. With an image size less than 40 km×40 km, their applications have been limited.

Availability of data

Data from China's remote sensing systems are available from multiple institutions with little standardization of ordering delivery or product specification, not to mention data policy relating to subsequent sharing.

Data acquisition and processing institutions

China first established its own satellite ground receiving and processing systems in the 1980s for foreign satellites. There are now corresponding institutes for each of the different classes of satellites under different government departments. Their main mission is to receive, process, archive, and distribute data from various remote sensing satellites and also to carry out research and development to improve data reception, processing, archiving, and distribution. China's operational remote sensing satellite data service network consists of the China Remote Sensing Satellite Ground Station (China RSGS), the China Center for Resource Satellite Data and Applications (CRESDA), the National Satellite Meteorological Center (NSMC), and the National Satellite Ocean Application Service (NSOAS). The existence of these centers has not yet led to the widespread distribution of data products within or outside China.

Data policy evolution

In China, there is no unified set of policies established by the central government regarding the ownership and rights of distribution of Chinese satellite data. ‘For a long time the attitude that public scientific data in China are the exclusive private possessions of individuals and departments has made the development of scientific data exchange and sharing difficult’ (Cheng 2006). Each of the above institutions belongs to different governmental departments and has its own managerial organizations, such as the Chinese Academy of Sciences, the China National Space Administration, the China Meteorological Administration, the State Oceanic Administration, the Ministry of Environmental Protection, and the Ministry of Civil Affairs. Some operational institutions of satellites belong to more than one department and each administrative department has its own data policy. Some organizations have quite liberal sharing policies whereas others are much more restrictive; this makes data acquisition very difficult even for domestic users within China.

In order to raise the awareness of scientific data sharing and to enhance its effectiveness, in 2002, the State Council of China authorized the Ministry of Science and Technology (MOST) to initiate a pilot project of the national science and technology infrastructure platform in coordination with 16 other ministries and departments to promote data sharing in China (Cheng 2006). This was a milestone in China for data sharing in putting forward the view to Chinese institutions and departments that public scientific data are national resources and that no organization or individual should be allowed to keep them privately. Although understood as a principle, free and open sharing of remotely sensed data is still uncommon in China.

Current status of data sharing in China

With the coordination from high-level governmental departments, data sharing between organizations within China has started to improve. China is actively participating in many international activities such as the Committee on Earth Observing Satellites (CEOS) and the Global Earth Observation System of Systems (GEOSS), who encourage data sharing. GEOSS pursues a long-term strategic plan made by the intergovernmental Group on Earth Observations (GEO), which is the largest inter-governmental organization in the earth observation area. It includes nearly 90 Member States and 67 Participating Organizations who are striving to integrate earth observations so that informed decisions can be made across nine Societal Benefit Areas including agriculture, biodiversity, climate, ecosystems, energy, disasters, health, water, and weather. Therefore, China is developing a more open policy by sharing Chinese satellite data not only between users within China but also with users from all over the world. Particularly noteworthy in this regard was opening and sharing CBERS data, the high-level data access initiative announced by Gang Wan, the Chinese Minister of Science and Technology at the Cape Town Ministerial Summit in November 2007 on GEO. Furthermore, the State Oceanic Administration (SOA) announced open access to data from the oceanic surveying satellite Haiyang-2 in 2012 (Geospatial World 2012).

By introducing open data policy from overseas, China has gradually expanded the range of sharing of different types of satellite data, from the FY series data to CBERS, HJ, and the oceanic surveying satellite Haiyang-2, making them open to domestic users and in some cases to foreign users. Generally speaking, data policy is becoming more open. Originally all data access was closed and only can be distributed to other domestic departments with the exceptional approval from its overseeing administrative departments and it was still charged sometimes. Now most data is open to domestic users though with certain conditions such as registering and signing agreements about usage and sharing. Today more and more data are open to the world for scientific research on global change and environmental monitoring. But most of them are only accessible with certain condition and constraints. Below we summarize some of the availability and restrictions associated with Chinese satellite data.

Earth resources satellites and environmental monitoring satellites (CBERS and HJ)

Data are open to domestic users under certain conditions. For example, CBERS-2B data can be distributed freely under the condition of signing a confidentiality agreement and the joint development of satellite data promotion agreements between users and CRESDA. For some data in specific region (e.g. Africa), data can be distributed overseas based on approval from CRESDA's overseeing administrative departments such as the National Defence Science and Technology Industry Council and the Ministry of Industry and Information Technology of the P.R. China. The HJ operational environmental and disaster monitoring satellites are operated by the Satellite Environment Center (SEC), which is a subsidiary institution of the Ministry of Environment Protection of China (MOEP). But the SEC will be supervised jointly by both MOEP and the Ministry of Civil Affairs of China (MOCA) in terms of operation and management of HJ satellite data (Wang 2012). SEC has to follow the data policies of both MOEP and MOCA. With multi-operational agencies and administrative departments, it becomes even more complex to get approvals for sharing HJ satellite data with international users.

Meteorological satellites and marine satellites (FY and HY)

In principle, data are open to domestic users and international users after registration online. Currently, FENGYUNCast, a national-level satellite remote sensing data sharing platform, is operated by NSMC. This data sharing system contains a ground application system composed of four receiving stations and one data-processing center obtaining polar-orbiting meteorological satellite data of the Chinese FY series, American NOAA series, EOS/MODIS series, the European MSG (Meteosat Second Generation), as well as China's geostationary meteorological satellite data from FY-2C/D and the Japanese MTSAT (NSMC 2009). FENGYUNCast's products are openly shared with both domestic users and international organizations such as WMO and GEO. But downloading the huge amounts of data will require the development of cooperation between Chinese providers and overseas users. NSOAS is under the jurisdiction of the State Oceanic Administration, mainly responsible for development of ocean satellite series and satellite ocean applications (NSOAS 2012). So far there is no online operational system to provide data to the public as yet. Data are distributed strictly under individual bilateral agreements with the provider.

Small satellites (Hangtian Tsinghua-1 and Beijing-1)

These satellites are owned by universities and private companies and are for scientific research and experiments and commercialization. Hence a commercial model is applied for acquisition of such data. As for most such businesses, the data are open to any users but will require charging and provision under contract especially relating to sharing and reproduction.

Potential data from China

Overall, China has significant potential to contribute to digital earth through its remote sensing satellite for global environment monitoring and protection. But open access to data is essential to fulfill this potential. Considerable contributions can be made by China's land satellite observation systems, including the ‘ZY’ earth resource satellite series and the small satellite constellation known as the HJ series of satellites for environment and disaster monitoring and forecasting. These satellites series can produce moderate and high resolution remote sensing data, based on which high resolution and ultra-high resolution systems can be developed. With the less than two days revisit time, these remote sensing satellites can provide updated information in a timely and regular fashion. Satellites equipped with multispectral and hyperspectral sensors can cover the full spectral range from visible to microwave wavelengths (Xu 2012). Sensors in visible or near-infrared can provide data for either macro-scale resource surveys such as monitoring of land cover, land use, land use change, agriculture and forestry, or special applications, which require a wealth of spectral information such as precision agriculture, mineral exploration, and vegetation growth monitoring. Sensors in the infrared range can provide data in monitoring urban heat, forest fires, and more generally land surface temperatures. Sensors in the microwave range with the strong cloud penetrability can provide key data for disaster monitoring. Consequently, a stable and reliable data source for global earth observation and environment monitoring is available from China if its terrestrial satellite data can be made available openly to the world. This will also be helpful for high-level decision makers to deal with both regional and global environmental issues.

The GEOSS plan emphasizes the need for earth observation to be directed toward benefiting specific societal benefit areas. To achieve this goal will require a coordinated international effort based on comprehensive data sharing. GEOSS also emphasizes the need to use remote sensing observations for multiple purposes; to develop end-to-end information systems; to facilitate data access; and to establish comprehensive data policies with the agreement of GEOSS Data Sharing Principles of being full and open exchange of data, metadata, and products shared within GEOSS. All shared data, metadata, and products will be made available with minimum time delay and at minimum cost, and all shared data, metadata, and products are free of charge or no more than cost of reproduction will be encouraged for research and education (GEOSS 2009).

Improving access to data: populating digital earth

The World Summit on Sustainable Development held in Johannesburg in 2002 emphasized the need for improved coordinated observations of the earth. In response, the first ministerial summit on earth observation was held on 31 July 2003 in the capital of USA, Washington DC, where 33 countries, the European Union, and 21 senior officials of international organizations involved in earth observation attended the meeting. After two years of efforts, the 10-year implementation plan of the Global Earth Observation System of Systems (GEOSS) was approved at the 3rd Earth Observation Ministerial Summit Meeting on 17 February 2005 in Brussels. The summit decided to formally set up an inter-governmental organization called the Group for Earth Observation (GEO). GEO is responsible for the coordination and implementation of GEOSS, which is now supported by over 80 countries, the European Union, and more than 60 international organizations. Without doubt, China's involvement in GEOSS has the potential to bring considerable benefits. Adopting an open international data-sharing policy would make China a major internationally recognized contributor to GEOSS and ultimately to the sustainability of the earth.

GEO recognized that comprehensive sustained observation and understanding of the earth system would help improve and expand the global capacity to achieve sustainable development. The purpose of ‘GEO 2012–2015 Work Plan’ is to promote and coordinate surface-based and space-based observing systems to provide long-term continuous observations of all components of the earth system (atmosphere, ocean, terrestrial, ice, solid earth) and ensure that the Earth and its physical processes are monitored globally across spatial and temporal scales (GEO 2011). Realizing that data sharing will be a key element to fulfill this purpose, GEO set up aims to provide a shared, easily accessible, timely, sustained stream of comprehensive data of documented quality, as well as metadata and information products, for informed decision-making. This should occur by 2015 through preparation of and access to global and regional information encompassing and data availability.

In order to promote data sharing and availability, GEO formulated a set of data-sharing principles. GEOSS asked CODATA, a subsidiary body of the International Council for Science, to review previous data policies and make recommendations that were used to formulate the GEOSS Data Principles agreed to at the 3rd Earth Observation Summit in Brussels as part of the GEOSS 10 Year Implementation Plan (GEOSS 2009).

1. There will be full and open exchange of data, metadata, and products shared within GEOSS, recognizing relevant international instruments and national policies and legislation.

2. All shared data, metadata, and products will be made available with minimum time delay and at minimum cost.

3. All shared data, metadata, and products being free of charge or no more than cost of reproduction will be encouraged for research and education.

The functional results of this are expected to be demonstrated by the open, reliable, timely, consistent, and free access to a core set of essential environmental observations and information products, supported by adequate metadata, by users across all GEOSS Societal Benefit Areas in accordance with GEOSS Data Sharing Principles (GEOSS Strategic Targets 2009).

But note that the qualifications with respect to ‘recognizing relevant, national policies and legislation’ provides no binding commitments on nations to have open data-sharing policies and even the reference to ‘no more than the cost of reproduction’, although sounding very reasonable in principle, can impose unnecessarily restrictive policies (Section 1.2: Data Management, Page 8/19).

It is worth recalling that it is now over 20 years since the US government issued the data sharing principles known as the Bromley Principles, which placed no restrictions on any data sets used for global change research. With the exception of Landsat data for a period, this has allowed users worldwide unfettered access to all US civilian satellite remote sensing data. The Bromley Principles arguably remain the simplest elegant and helpful data-sharing principles ever derived. Universal adoption of them for remotely sensed data, including by China, would contribute enormously to the successful population of digital earth (Table 3).

Table 3. The Bromley Principles (Bromley 1991).

The Global Change Research Program requires an early and continuing commitment to the establishment, maintenance, validation, description, accessibility, and distribution of high-quality, long-term data sets.

Full and open sharing of the full suite of global data sets for all global change researchers is a fundamental objective.

Preservation of all data needed for long-term global change research is required. For each and every global change data parameter, there should be at least one explicitly designated archive. Procedures and criteria for setting priorities for data acquisition, retention, and purging should be developed by participating agencies, both nationally and internationally. A clearinghouse process should be established to prevent the purging and loss of important data sets.

Data archives must include easily accessible information about the data holdings, including quality assessments, supporting ancillary information, and guidance and aids for locating and obtaining the data.

National and international standards should be used to the greatest extent possible for media and for processing and communication of global data sets.

Data should be provided at the lowest possible cost to global change researchers in the interest of full and open access to data. This cost should, as a first principle, be no more than the marginal cost of filling a specific user request. Agencies should act to streamline administrative arrangements for exchanging data among researchers.

For those programs in which selected principal investigators have initial periods of exclusive data use, data should be made openly available as soon as they become widely useful. In each case, the funding agency should explicitly define the duration of any exclusive use period.

Examples of the benefits of open sharing

Landsat class data

The first large-scale free distribution of Landsat data was made by the Global Land Cover Facility at the University of Maryland, which was set up in 2001 as part of the Earth System Information Partnership funded by NASA. Landsat data were first made available at no cost in 2003. Downloads quickly rose to between 20,000 and 25,000 scenes per month at a time when the total number of scenes being purchased through the USGS was less than 20,000 scenes per year. The Landsat data were based on the three global collections made for 1975, 1990, and 2000 epochs. Subsequently global collections have been made available for 2005 and 2010. As we will discuss later, the large number of scenes downloaded was also a result of standardization of formats.

In January 2008, NASA and the United States Geological Survey (USGS) implemented a free Landsat Data Distribution Policy (Roy et al. 2011). Currently, newly acquired Landsat 5 and 7 data with <30% cloud cover are automatically processed and placed on-line for immediate downloading. All other Landsat scenes (over 2million) are available at no charge via an on-demand ordering and downloading capability. Shortly after initiating access, in January 2009 nearly 73,000 scenes were downloaded an average of almost 2400 scenes per day. As of 1 September 2012, nine million images had been downloaded from USGS in less than four years from the start of free downloading (UN-SPIDER 2012).

CBERS data were initially only made available through an ordering system, but in 1 April 2006 all CBERS-2 data were made available for free download by domestic users. At the time a total of 20,000 images had been distributed in the previous five years. Downloads leaped to more than 10,000 per month once open access to Chinese users occurred. Similarly, a very large increase in data volume was observed when the Institute for Geographical Sciences and Natural Resources Research made some of their database available in 2005. There was an almost 10-fold increase in the data products delivered.

NASA's EOS products

The strongest argument for the free and open distribution of data is that provided by NASA's Earth Observing System Data and Information System (EOS-DIS). Table 4 shows the huge volumes of data added and accessed every day during a year from 1 October 2011 to 30 September 2012 (corresponding to the US government's financial year).

Table 4. EOS-DIS metrics (Chang et al. 2013).

EOSDIS FY 2012 Metrics (1 October 2011 to 30 September 2012)
Unique Data Sets	6866
Distinct Users of EOSDIS Data and Services	1.5 M
Web Site Visits	2.6 M
Average Archive Growth	5.4 TB/day
Total Archive Volume	7.4 PB
End User Distribution Products	636 M
End User Average Distribution Volume	17.1 TB/day

Data distribution is freely available to all users including those users who are outside of the USA as shown in Table 5. Note that China only follows the USA in the number of distinct users and these are twice the number of the 3rd ranked country, Germany.

Table 5. US-EOS data users – top 20 countries (Chang et al. 2013).

Ranking	Country	Users
1	USA	141,916
2	China	38,854
3	Germany	19,180
4	France	19,161
5	United Kingdom	18,022
6	India	16,236
7	Canada	13,399
8	Russia	12,262
9	Brazil	12,037
10	Japan	8633
11	Australia	8239
12	Spain	6826
13	Italy	6680
14	Guatemala	5650
15	Indonesia	5019
16	Iran	4878
17	Netherlands	4344
18	Poland	4263
19	Mexico	4244
20	Argentina	3694

The considerable contribution of data sharing can also be demonstrated in the improving of the scientific research. MODIS data provided by NASA have been freely open to global users. Since 2000, MODIS data have been widely used to study the Earth's climate system. ‘Web of Science shows that more than 6500 scientific papers published using MODIS data that have been cited over 65,000 times as of 30 September 2011’ (Tucker and Yager 2011).

Commercial satellite data

Very recently, some of commercial data have been shared through government in the USA. The National Geospatial-Intelligence Agency has contracted United States commercial remote sensing companies GeoEye and Digital Globe to provide high resolution (<1 m) commercial satellite data (Neigh, Masek, and Nickeson 2013). The purpose of sharing data is to support federal/state government agencies and those projects/people who support government interests. Currently, some free data are already available through the US Geological Survey and NASA Earth Science community and programs such as the Land-Cover Land-Use Change program.

Fallacies about the need for charging for governmental remote sensing data

Despite the obvious benefits of making government-funded remote sensing data, freely available arguments persist in arguing that users should be made to pay (in the case of private companies, providing their business model may require payment for them to be viable). Here are some of the most common arguments and their rebuttals.

If people want to use remotely sensed data then they should pay

This statement seems reasonable, but really, it is not. The citizens of countries have by definition already paid for the remote sensing systems. Moreover, the driving force for most environmental remote sensing data is scientific or policy driven for use by governments. Wider use of the data leads to a better return on the investments made.

Making data available has an incremental cost, which should be borne by the user

To make data readily available to government users, by far the most efficient way is to make them available on-line. Making them available to all other users will result nowadays in a tiny incremental cost. Indeed the main cost involved in charging for data may well be the charging itself in the billing and receipts system itself. Moreover, any resources raised are likely to be a tiny fraction of the total cost of the system.

There is a commercial future for all environmental remote sensing data

Clearly commercial remote sensing has in some cases been successful, but there is no evidence yet that this is the case for mid and coarser resolution data for terrestrial application (Williamson 1997). In the case of very fine resolution data (<5m) from US companies, only the large contracts issued by the US government currently make these ventures viable. Moreover, making data freely and readily available has the potential to stimulate value-added enterprises focusing on products and services.

A restrictive data policy is acceptable because remote sensing data are made available free to scientists

This begs the question as to why should scientists have preferential access rather than other users such as those working in developing countries to alleviate poverty.

Other barriers to the use of remote sensing data

Various other barriers exist, limiting user's access. For example, sometimes principle investigators have been given inordinately long periods of exclusive use justified by the need to ensure high-quality products are distributed. A better approach is to make sure that the products meet minimum standards of quality and then make them available with appropriate guidance on quality. Evidence from the US EOS program clearly shows that multiple sets of reprocessing are needed: MODIS data sets are about to have Collection 6 released, which represents an original data release and then 5 sets of completely reprocessed data. Waiting for perfection makes no sense.

Another process that restricts use is to ask users to explain their use of the data as has been requested by the European Space Agency (ESA) for example. New users are readily put off by such requests and it is a pointless exercise given the very low marginal costs of making the data freely available.

A pernicious restriction to greater use of data is to make data sets available in user-unfriendly formats. As originally delivered, MODIS products were in an integerized sinusoidal projection (ISIN) unreadable by any standard image processing or Geographic Information System (GIS) and to generate global or even regional images required the stitching together of very large tiles. In principle, though not necessarily in practice, all space agencies should make data available in formats that allow ready ingestion and for terrestrial users this includes immediate ingestion into GIS's. For example, the University of Maryland's Global land Cover Facility has over a decade made all of its Landsat data available in a GeoTIFF file format, using either Geographic coordinates or Universal Transverse Mercator (UTM) coordinates projections with a World Geodetic System 1984 (WGS84) datum and GNU zip compression. The USGS has adopted very similar standards in recent years.

The decision of the USGS to make all of its Landsat data available in an ortho-rectified form is also very important to reduce misregistration in more rugged areas and is a standard that should be adopted by all space agencies. Wherever possible, atmospheric correction of data should always be performed so that the images represent the surface signal but not the one contaminated by the variable atmospheric signal. For Landsat TM data, this has been shown to be feasible (Feng et al. 2013). For other sensors with a more limited set of bands and without near-simultaneous overpass with the excellently characterized MODIS sensor, this will be more challenging.

Many ordering and delivery systems are far from optimal. It is making it time consuming to download data especially when large numbers of images need to be ordered. Directory catalogs and ordering mechanisms remain very variable even from the same institution, such as the USGS EROS data center.

Yet another barrier is the so-called ‘Valley of Death’, which refers to the key stage of transferring R&D capabilities to operational implementation. For technology investments, the transitions from development to implementation are frequently difficult. Although it is a general problem, this has been noted as a particular one with respect to remote sensing satellites (Board on Atmospheric Sciences and Climate NRC 2000). The issue here is that unless there is a regular supply of data and products, operational users will be reluctant to adopt them for their scientific or operational purposes. Successful transitions from R&D to operational implementation requires consideration of several issues including:

• understanding of the importance (and risks) of the transition,

• development and maintenance of appropriate transition plans,

• adequate resource provision, and

• continuous feedback (in both directions) between the R&D and operational activities.

In terms of remote sensing, this is a huge problem in ensuring the long-term continuity of data sets. Ensuring that satellites are provided to ensure long-term records is often difficult to achieve – witness the challenges in making the US Landsat system an operational one, where every individual satellite has had to be justified separately.

Recommendations for improved use of Chinese data to populate Digital Earth

As the largest country in Asia, China is significant in providing remote sensing data due to its distinctive geographical location of sitting in the East Asia Monsoon Belt that experiences the largest variation of earth environment, the highest mountain of Himalaya, and the 9.6 million square kilometers acreage of whole country. Consequently, the Chinese Satellites and the data resources recorded from in situ, aircraft, and satellite networks are essential to GEO and GEOSS for global environment monitoring (Li, Townshend, and Wu 2010).

Proposed data principles

There are many other key principles (see above in section 5) that could be adopted by China especially those relating to the creation of high-quality, consistent, long-term records of global change. Various attempts have been made to establish such data principles like the Bromley principles mentioned above. Note that some items are most relevant to the discussion of availability and costs of data. As originally formulated, these principles were applied to global change research, but US government subsequently adopted them for many other applications. Refinements and development of these principles have been developed by many organizations, for example, the Global Climate Observing System (GCOS), the Integrated Global Observing Strategy Partnership (IGOS-P), and more recently by the intergovernmental Group on Earth Observations (GEO).

It is recommended that consideration is given to the adoption of further data principles to ensure that the maximum benefit is received by the Chinese people for the considerable investments made in remote sensing. Also it is recommended that consideration will be given to the participation of space agencies including both domestic space agencies and international space agencies. Context as following can be included and emphasized in data policies and principles in China:

1. Encourage the growing commitment of space agencies. This is a substantial part to ensure there are sufficient resources for ‘validation’ of products provided by space agencies.

2. Establish an international-standard coordinated assessment, which is independent of the agencies producing data and products. This will lead to friendly used format of data to worldwide users.

3. As a key member of GEO, CEOS, and IGOS-P, China is encouraged to adopt these international organizations’ data policies leading to free and open access of remotely sensed data. As users representing a wide range of scientific and application areas, we are seriously concerned that without the implementation of such a common data policy the goals of GEOSS will be seriously compromised.

4. Efficient operational systems are essential to increase use of data through actions such as improved global acquisition strategies, delivery of products with standard user-friendly formats by all agencies, and China should participate in such activities.

Thoughts on implementation principles

No matter how well the data policies and principles are made, there are always technical challenges for any country or international organization in the implementation stage. Yunck et al. (2006) listed the guiding themes that are needed to be considered in operational implementation of sharing data. These guiding themes include decentralization; semantic integration; peer-to-peer exchange; machine-to-machine, automated workflows; distributed execution; dynamic load balancing; grid and web services; multi-scale integration; as well as plug-and-play software. Hence, policies relating to the distribution and availability of remote sensing data are highly recommended to be included into data policies and becoming part of the operational policy to navigate implementation of sharing data. A very good example of conducting data sharing effectively can be found in Integrated Global Observation Strategy (IGOS) Data and Information Systems and Service (DISS) principles (Townshend et al. 2008). Based on the inspiration from IGOS and DISS, we recommend some points to be considered as implementation guidelines:

1. To build up a user-orientated system that can maximize severing worldwide users including potential users in the future. This system can be a network composed of data collection, data processing, data distributing, product generation, and archival activities.

2. To specify the responsibilities of agencies in each aspect of the system including data providing, processing, generation, and distribution. Self-regulation system is recommended as a complementary managerial system to data-sharing policy and principle.

3. To build up an advanced technical system that has the capability and flexibility of adjustment and coordination between currently existing data sharing systems and potential expanded data sharing systems in the future. To bear the awareness that potential locations, other than data centers, can be coordinated as a part of the system to be accessed for data products and services when designing implementation policy.

4. To make sure of providing user-friendly data products that are easy to be accessed, easy to be used, and validated. If possible, a rapid and convenient access to data, products, and services should be available in an easy way for users who are with different technical skills and purpose of using remote sensing data.

5. To have tool kits available in the operational system. This will help users to cope with potential challenges in the future, when new transition to operational systems happens. Also, evaluation of the impacts to the users due to potential changes in instrumentation and other processing techniques need to be factored into the planning of new systems.

6. To implement sufficiently under guidance of a reasonable policy and principle. A technical team with knowledge and expertise in data information system and services is another essential factor to strength the operation and sustainability. Therefore, to build up an experienced managerial team, technical team and network is the functional driven force to conduct data sharing policies and principles.

Over all, data sharing is a comprehensive system, which includes data collection, processing, generation, distribution, and monitoring. Consequently, a practical data policy and principle should have considerations on every aspect, which it might encounter for operating of data sharing. A comprehensive policy covers the completed stages including designing, operation, and feedback. Implementation policy should consider phases of pre-operation, operation, and post-operation of data sharing. Having a national data system that can avoid duplicate or replicated efforts for specific purposes can promote and encourage the work of related institutions in support of national or international objectives. It will benefit not only China but also the world if a national data-sharing system can be integrated into the world data sharing system.

Acknowledgements

We gratefully acknowledge the comments from anonymous reviewers, IJDE editors, and Xiaopeng Song, one of our colleagues in the Department of Geographical Sciences at the University of Maryland.