:Draft:Overhead Image Revolution
{{AfC submission|t||ts=20250417093354|u=Ziliujiang123|ns=118|demo=}}
1.Introduction
Since the dawn of the 21st century, the human demand for Earth observation has grown exponentially. From monitoring global climate change and forecasting natural disasters to managing resources, environment, urban planning, agricultural production, infrastructure development, and even ensuring national security, timely, accurate, and comprehensive Earth observation data has become indispensable [1]. Traditional Earth observation methods predominantly rely on satellite platforms operating in low Earth orbit (LEO, 500-800km altitude) or geostationary orbit (GEO, 36,000km altitude). While these systems have provided invaluable data, they inherently face limitations in spatial resolution, revisit frequency, and application agility.
The evolution of Earth observation, marked by milestones such as the launch of Landsat-1 in 1972 [2], the establishment of NASA's Earth Observing System (EOS) in 1991 [2], and the ongoing efforts of the Group on Earth Observations (GEO) to coordinate the Global Earth Observation System of Systems (GEOSS) [3], has seen gradual progress in technology adoption, industry penetration, and public awareness. The "Economic Valuation of Landsat and Landsat Next 2023" report [4] highlights the significant economic value generated by Landsat, estimating $25 billion in direct benefits and over $2 billion in indirect benefits for the US in 2023, with Landsat Next projected to generate $33 billion annually. However, the spatial resolution of up to 15m and temporal resolution of "days" for Landsat data continue to draw skepticism from Earth observation data users and industry observers regarding its true value in addressing increasingly granular needs.
Concurrently, the rise of "NewSpace" [5], the explosion of artificial intelligence, and the development of the "low-altitude economy" are fueling significant interest and anticipation in the Earth observation sector. The availability of satellite and aerial imagery with resolutions as high as 15cm, along with large-scale (1:5000) terrain models on platforms like Google Earth [6], underscores this trend. Companies like Maxar are achieving production rates of over one million square kilometers of 15cm resolution and three million square kilometers of 30cm resolution imagery daily, distributing these data products and services globally through their Amazon Web Services-based Maxar Geospatial Platform (MGP) [7]. Preligens' SSS service [8] offers daily monitoring of tens of thousands of strategic locations worldwide. Albedo [9] is poised to distribute commercial satellite imagery with a top resolution of 10cm from very low Earth orbit. Near Space Labs [10] is already commercially distributing 7cm overhead imagery covering the entire United States quarterly using stratospheric balloon-borne robots. Furthermore, the OpenAerialMap website, under a CC-BY 4.0 license [11], and the OpenStreetMap Humanitarian Team's Open Imagery Network (OIN) nodes provide free access to drone imagery with resolutions ranging from 2 to 5cm.
These developments signify a growing recognition and rapid expansion of the immense value inherent in higher-resolution, nadir-viewing remote sensing imagery [12]. This trend is highly likely to trigger a fundamental revolution in Earth observation technology, which we term the "Overhead Image Revolution" (OIR).
2.Concept
The "Overhead Image Revolution" refers to the acquisition of high spatial resolution, high temporal resolution, high spectral resolution, and high-precision Earth observation, monitoring, and detection data from remote sensors operating in five primary flight domains: low orbit (450-2000km altitude), very low orbit (160-300km) [14], stratosphere (20-50km) [15], aerial layer (3-20km), and low altitude (0-3km). The resulting data is transformed into ubiquitous geospatial intelligence, serving various aspects of human life, production, scientific research, commerce, entertainment, and defense. This revolution is anticipated to drive comprehensive social progress and even civilizational upgrades within the next 10 to 20 years.
= 2.1 Characteristics =
== 2.1.1 Emphasis on High Resolution and High Accuracy ==
The "Overhead Image Revolution," as a novel concept, has a clear value proposition: "High Resolution + High Timeliness + High Accuracy = High Value." It is important to note that "high value" here emphasizes the direct problem-solving capability and the resulting direct value, such as assessing the area and potential of rooftop distributed photovoltaics in urban environments [16]. This does not negate the value and contributions of traditional Earth observation systems, which operate at scales from global to regional and focus on the atmosphere, oceans, and land, primarily serving global public interests and sustainable development goals.
The definition of high resolution, high timeliness, and high accuracy is determined by the problem-solving needs in high-value application scenarios. Taking high spatial resolution in optical multispectral remote sensing as an example: low orbit is suitable for achieving around 30cm resolution, very low orbit around 10cm, stratosphere around 5-10cm, aerial layer around 5cm, and low altitude around 2-5cm. Achieving such high spatial resolution at scale holds the potential to address numerous high-value application scenarios for Earth observation users, including digital twin cities, 3D reality Earth, land resource monitoring, precision agriculture, intelligent transportation, smart energy, and national defense.
== 2.1.2 Emphasis on Vertical Integration of the "Earth Observation System of Systems" ==
As mentioned earlier, GEO is coordinating the development of GEOSS, which primarily focuses on the horizontal integration of global ocean, land, and climate observations, with the main objective of explaining and predicting macroscopic phenomena in the atmosphere, oceans, and land.
The "Overhead Image Revolution" does not exclude the horizontal integration of GEOSS but places a greater emphasis on the vertical integration of Earth observation systems across the five primary flight domains (low orbit, very low orbit, stratosphere, aerial layer, and low altitude) within GEOSS. This vertical integration aims to answer and solve "5W+2H" questions [17] in specific scenarios, a capability that is currently a "shortcoming" of the existing GEOSS and a key to its future development and widespread adoption.
== 2.1.3 Emphasis on Deep Integration with Artificial Intelligence ==
The "Overhead Image Revolution" emphasizes the widespread application and service of geospatial intelligence. The continuous acquisition of overhead imagery and its deep integration with artificial intelligence for processing and interpretation are inherent requirements. According to Maxar experts, by 2025, interpreting and analyzing the daily global acquisition of Earth observation imagery would require at least 8 million intelligence analysts if relying solely on manual efforts, which significantly exceeds available human resources. The realization of the "Overhead Image Revolution" necessitates advanced artificial intelligence capabilities for multi-modal data processing and interpretation. Conversely, the continuous stream of overhead imagery will help alleviate the "data wall" [18] problem in artificial intelligence and even aid in identifying AI "hallucinations" and "misinformation."
References
{{reflist}}[1] Li Deren, Zhang Guo, Jiang Yonghua, et al. Opportunities and Challenges of Geospatial Informatics under the Perspective of Big Data [J]. Big Data, 2022, 8(02):3-14.
[2] NASA, Nasa’s earth observation system project science office,
[3] GEOSS, About GEOSS,
[4] USGS, Economic Valuation of Landsat and Landsat Next 2023,
[5] G. Martin, NewSpace: The Emerging Commercial Space Industry,
[6] Google Earth,
[7] Maxar Technologies , Maxar Geospatial Platform(MGP) ,
[8] Safran, Geospatial Intelligence AI,
[9] Satellite Imaging Corporation, Albedo Satellite Constellation,
[10] Businesswire, Near Space Labs Announces Nationwide Deployment of Stratospheric Robots Featuring New Industry-Leading 7cm Resolution Aerial Imagery,
[11] OpenAerialMap,
[12] Wikipedia, Overhead Imagery Research Data Set,
[13] Zhang Ming. Development and Challenges of Low Orbit Satellite Systems [J]. China Radio, 2019,(03):56-57.
[14] Lü Jiuming, Lu Jiangong, Diao Jingjing, et al. Development Status and Application of Very Low Orbit Satellite Technology [J]. National Defense Science and Technology, 2020, 41(01):33-37. DOI:10.13943/j.issn1671-4547.2020.01.09.
[15] Yu Chunrui, Qiao Kai, Liu Dongxu. Basic Problems in the Overall Design of Stratospheric Airships [J]. Journal of Space Science, 2022, 42(05):927-932.
[16] Hou Jiang, etc. Geospatial assessment of rooftop solar photovoltaic potential using multi-source remote sensing data, Energy and AI, Vol. 10, 2022, 100185, ISSN 2666-5468,
[17] HEFLO, The 5W2H Method: learn how to create a simple action plan,
[18] R. Shrivastava, The Prompt: What Happens When We Hit The ‘Data Wall’?