Big Geo Data Research Articles

Digital Coal Geological Map

From 1974 to 1994, the Korea Institute of Geoscience and Mineral Resources (KIGAM) systematically prepared and published relatively precise coal cell geology maps of 1:10,000 or 1:25,000 scale for major coal fields across the country. Such a coal cell geology map includes information about the coal seams as well as the geology of the coal field area, so it can be used as an important basic data for coal development. In this paper, the current state of the coal geology map, which was digitized into a spatial DB using GIS, was introduced. This digital coal geological map can be downloaded free of charge from the Geo-Big Data Open Platform (data.kigam.re.kr) and the Environmental Big Data Platform (www.bigdata-environment.kr).

The Premium of Public Perceived Greenery: A Framework Using Multiscale GWR and Deep Learning.

Population agglomeration and real estate development encroach on public green spaces, threatening human settlement equity and perceptual experience. Perceived greenery is a vital interface for residents to interact with the urban eco-environment. Nevertheless, the economic premiums and spatial scale of such greenery have not been fully studied because a comprehensive quantitative framework is difficult to obtain. Here, taking advantage of big geodata and deep learning to quantify public perceived greenery, we integrate a multiscale GWR (MGWR) and a hedonic price model (HPM) and propose an analytic framework to explore the premium of perceived greenery and its spatial pattern at the neighborhood scale. Our empirical study in Beijing demonstrated that (1) MGWR-based HPM can lead to good performance and increase understanding of the spatial premium effect of perceived greenery; (2) for every 1% increase in neighborhood-level perceived greenery, economic premiums increase by 4.1% (115,862 RMB) on average; and (3) the premium of perceived greenery is spatially imbalanced and linearly decreases with location, which is caused by Beijing’s monocentric development pattern. Our framework provides analytical tools for measuring and mapping the capitalization of perceived greenery. Furthermore, the empirical results can provide positive implications for establishing equitable housing policies and livable neighborhoods.

Predicting wildfire burns from big geodata using deep learning

Wildfire continues to be a major environmental problem in the world. To help land and fire management agencies manage and mitigate wildfire-related risks, we need to develop tools for mapping those risks. Big geodata—in the form of remotely sensed images, ground-based sensor observations, and topographical datasets—can help us characterize the dynamics of wildfire related events. In this study, we design a deep fully convolutional network, called AllConvNet, to produce daily maps of the probability of a wildfire burn over the next 7 days. We applied it to burns in Victoria, Australia for the period of 2006–2017. Fifteen factors that were extracted from six different datasets and resulted into 29 quantitative features, were selected as input to the network. We compared it with three baseline methods: SegNet, multilayer perceptron, and logistic regression. AllConvNet outperforms the other three baseline methods in four of the six quantitative metrics considered. AllConvNet and SegNet provide smoother and more regularized predicted maps, with SegNet providing greater sensitivity in dificriminating less wildfire-prone locations. Input feature statistical importance was measured for all the networks and compared against logistic regression coefficients. Total precipitation, lightning flash density, and land surface temperature occur to be consistently highly weighted by all models while terrain aspect components, wind direction components, certain land cover classes (such as crop field and woodland), and distance from power lines are ranked on the lower end. We conclude that wild-fire burn prediction methods based on deep learning present quantitative and qualitative gains.

Construction of database and distribution map of mineral components for natural Good-Water in Korea

The database was established by conducting field surveys and water quality analysis of 877 water samples based on the national-scale various water sources, including 167 sampling sites from historically well-known hot springs, mineral waters, and spring waters (from August 2017 to June 2020). Spatial analysis using Kriging interpolation was performed to create a distribution map for 14 mineral components. The distribution maps were prepared with 50 m spatial resolution in raster type, and provided by Geo Bigdata Open Platform service of KIGAM. The spatial distribution of each mineral component showed some relationships with geological (lithology and fault) and regional characteristics, and could be used for drinking water and beverage industry (beverages, alcoholic beverages, teas, coffee, etc.) by providing the various information on the occurrence of natural Good-Water. We plan to continuously update information on Good-Water source to increase spatial resolution and secure its representativeness at each location.

Decoding the Street-Based Spatiality of Urban Gyms: Implications for Healthy City Planning

Many cities face health issues that result from ineffective urban planning strategies. The chances of doing exercises in sportive venues implicate public health and citizen quality of life. With the advent of the geo-big data era, it is crucial to explore the spatial pattern of sports facilities to reflect urban health issues. This study aims to decode the street-based spatiality of gyms (one prevailing type of sportive venues) from a comprehensive perspective by both geometric methods (i.e., segment streets) and topological analytics in the context of complexity science (i.e., complex network derived from the topology of natural streets). We found that: (1) gyms are spatially clustered and distributed unevenly; (2) community-to-gym walkability fits the power-law with a heavy-tailed distribution at the 10-min and 20-min temporal scales; (3) the model for the street connectivity and the multi-distance reachability of gyms is with high polynomial fitting goodness. This article is conducive to strategies-making of healthy city planning and the further optimization of urban spatial structure.

On increasing the productive time of drilling oil and gas wells using machine learning methods

The article is devoted to the development of a hybrid method for predicting and preventing the development of troubles in the process of drilling wells based on machine learning methods and modern neural network models. Troubles during the drilling process, such as filtrate leakoff; gas, oil and water shows and sticking, lead to an increase in unproductive time, i.e. time that is not technically necessary for well construction and is caused by various violations of the production process. Several different approaches have been considered, including based on the regression model for predicting the indicator function, which reflects an approach to a developing trouble, as well as anomaly extraction models built both on basic machine learning algorithms and using the neural network model of deep learning. Showing visualized examples of the work of the developed methods on simulation and real data. Intelligent analysis of Big Geodata from geological and technological measurement stations is based on well-proven machine learning algorithms. Based on these data, a neural network model was proposed to prevent troubles and emergencies during the construction of wells. The use of this method will minimize unproductive drilling time.

A Deep Learning Approach to the Detection of Gossans in the Canadian Arctic

Gossans are surficial deposits that form in host bedrock by the alteration of sulphides by acidic and oxidizing fluids. These deposits are typically a few meters to kilometers in size and they constitute important vectors to buried ore deposits. Hundreds of gossans have been mapped by field geologists in sparsely vegetated areas of the Canadian Arctic. However, due to Canada’s vast northern landmass, it is highly probable that many existing occurrences have been missed. In contrast, a variety of remote sensing data has been acquired in recent years, allowing for a broader survey of gossans from orbit. These include band ratioing or methods based on principal component analysis. Spectrally, the 809 gossans used in this study show no significant difference from randomly placed points on the Landsat 8 imageries. To overcome this major issue, we propose a deep learning method based on convolutional neural networks and relying on geo big data (Landsat-8, Arctic digital elevation model lithological maps) that can be used for the detection of gossans. Its application in different regions in the Canadian Arctic shows great promise, with precisions reaching 77%. This first order approach could provide a useful precursor tool to identify gossans prior to more detailed surveys using hyperspectral imaging.

A method to evaluate task-specific importance of spatio-temporal units based on explainable artificial intelligence

ABSTRACT Big geo-data are often aggregated according to spatio-temporal units for analyzing human activities and urban environments. Many applications categorize such data into groups and compare the characteristics across groups. The intergroup differences vary with spatio-temporal units, and the essential is to identify the spatio-temporal units with apparently different data characteristics. However, spatio-temporal dependence, data variety, and the complexity of tasks impede an effective unit assessment. Inspired by the applications to extract critical image components based on explainable artificial intelligence (XAI), we propose a spatio-temporal layer-wise relevance propagation method to assess spatio-temporal units as a general solution. The method organizes input data into an extensible three-dimensional tensor form. We provide two means of labeling the spatio-temporal tensor data for typical geographical applications, using temporally or spatially relevant information. Neural network training proceeds to extract the global and local characteristics of data for corresponding analytical tasks. Then the method propagates classification results backward into units as obtained task-specific importance. A case study with taxi trajectory data in Beijing validates the method. The results prove that the proposed method can evaluate the task-specific importance of spatio-temporal units with dependence. This study also attempts to discover task-related knowledge using XAI.

Understanding house price appreciation using multi-source big geo-data and machine learning

Understanding house price appreciation benefits place-based decision makings and real estate market analyses. Although large amounts of interests have been paid in the house price modeling, limited work has focused on evaluating the price appreciation rate. In this study, we propose a data-fusion framework to examine how well house price appreciation potentials can be predicted by combining multiple data sources. We used data sets including house structural attributes, house photos, locational amenities, street view images, transportation accessibility, visitor patterns, and socioeconomic attributes of neighborhoods to enrich our understanding of the real estate appreciation and its predictive modeling. As a case study, we investigate more than 20,000 houses in the Greater Boston Area, and discuss the spatial dependency of house price appreciations, influential variables and their relationships. In detail, we extract deep features from street view images and house photos using a deep learning model, merging features from multi-source data and modeling house price appreciation using machine learning models and the geographically weighted regression at two spatial scales: fine-scale point level and aggregated neighborhood level. Results show that the house price appreciation rate can be modeled with high accuracy using the proposed framework (R2=0.74 for gradient boosting machine at neighborhood-scale). We discovered that houses with low house prices and small house areas may have a higher house appreciation potential. Our results provide insights into how multi-source big geo-data can be employed in machine learning frameworks to characterize real estate price trends and help understand human settlements for policy-making.

Google Earth Engine for geo-big data applications: A meta-analysis and systematic review

Google Earth Engine (GEE) is a cloud-based geospatial processing platform for large-scale environmental monitoring and analysis. The free-to-use GEE platform provides access to (1) petabytes of publicly available remote sensing imagery and other ready-to-use products with an explorer web app; (2) high-speed parallel processing and machine learning algorithms using Google’s computational infrastructure; and (3) a library of Application Programming Interfaces (APIs) with development environments that support popular coding languages, such as JavaScript and Python. Together these core features enable users to discover, analyze and visualize geospatial big data in powerful ways without needing access to supercomputers or specialized coding expertise. The development of GEE has created much enthusiasm and engagement in the remote sensing and geospatial data science fields. Yet after a decade since GEE was launched, its impact on remote sensing and geospatial science has not been carefully explored. Thus, a systematic review of GEE that can provide readers with the “big picture” of the current status and general trends in GEE is needed. To this end, the decision was taken to perform a meta-analysis investigation of recent peer-reviewed GEE articles focusing on several features, including data, sensor type, study area, spatial resolution, application, strategy, and analytical methods. A total of 349 peer-reviewed articles published in 146 different journals between 2010 and October 2019 were reviewed. Publications and geographical distribution trends showed a broad spectrum of applications in environmental analyses at both regional and global scales. Remote sensing datasets were used in 90% of studies while 10% of the articles utilized ready-to-use products for analyses. Optical satellite imagery with medium spatial resolution, particularly Landsat data with an archive exceeding 40 years, has been used extensively. Linear regression and random forest were the most frequently used algorithms for satellite imagery processing. Among ready-to-use products, the normalized difference vegetation index (NDVI) was used in 27% of studies for vegetation, crop, land cover mapping and drought monitoring. The results of this study confirm that GEE has and continues to make substantive progress on global challenges involving process of geo-big data.

A Citizen-Centric Approach for the Improvement of Territorial Services Management

In the last decade, there has been a growing awareness that the involvement of citizens in decision making can produce an immediate and positive impact on actions to be taken, as they are the real owners of knowledge about the place where they live. By collecting and geolocating data through smartphones and the Internet, citizens in fact can help decision makers both create sharable spatio-temporal information about objects and phenomena and interpret territorial dynamics. However, although such a role has been definitely recognized, the lack of a homogeneous paradigm for structuring the sensing process, managing the geo big data produced and handling services makes it difficult to exploit such a potentiality. In this paper, we describe a citizen-centric approach conceived to build territorial knowledge useful to provide decision makers with a timely and reliable picture of the status of a given territory. In particular, a visual representation of geospatial knowledge is described, which summaries context-sensitive information about a territory and its citizens, thus improving the land monitoring tasks. An information system, SAFE, is finally presented, which consists of a Web and a mobile component to manage citizen supplied data to be integrated for building reliable dynamic scenarios.

Big geodata mining: Objective, connotations and research issues

The objective, connotations and research issues of big geodata mining were discussed to address its significance to geographical research in this paper. Big geodata may be categorized into two domains: big earth observation data and big human behavior data. A description of big geodata includes, in addition to the “5Vs” (volume, velocity, value, variety and veracity), a further five features, that is, granularity, scope, density, skewness and precision. Based on this approach, the essence of mining big geodata includes four aspects. First, flow space, where flow replaces points in traditional space, will become the new presentation form for big human behavior data. Second, the objectives for mining big geodata are the spatial patterns and the spatial relationships. Third, the spatiotemporal distributions of big geodata can be viewed as overlays of multiple geographic patterns and the characteristics of the data, namely heterogeneity and homogeneity, may change with scale. Fourth, data mining can be seen as a tool for discovery of geographic patterns and the patterns revealed may be attributed to human-land relationships. The big geodata mining methods may be categorized into two types in view of the mining objective, i.e., classification mining and relationship mining. Future research will be faced by a number of issues, including the aggregation and connection of big geodata, the effective evaluation of the mining results and the challenge for mining to reveal “non-trivial” knowledge.

A Scalable Synthesis of Multiple Models of Geo Big Data Interpretation

The paper proposes a scalable fuzzy approach for mapping the status of the environment integrating several distinct models exploiting geo big data. The process is structured into two phases: the first one can exploit products yielded by distinct models of remote sensing image interpretation defined in the scientific literature, and knowledge of domain experts, possibly ill-defined, for computing partial evidence of a phenomenon. The second phase integrates the partial evidence maps through a learning mechanism exploiting ground truth to compute a synthetic Environmental Status Indicator (ESI) map. The proposal resembles an ensemble approach with the difference that the aggregation is not necessarily consensual but can model a distinct decision attitude in between pessimistic and optimistic. It is scalable and can be implemented in a distributed processing framework, so as to make feasible ESI mapping in near real time to support land monitoring. It is exemplified to map the presence of standing water areas, indicator of water resources, agro-practices or natural hazard from remote sensing by considering different models.

Google Earth Engine Cloud Computing Platform for Remote Sensing Big Data Applications: A Comprehensive Review

Remote sensing (RS) systems have been collecting massive volumes of datasets for decades, managing and analyzing of which are not practical using common software packages and desktop computing resources. In this regard, Google has developed a cloud computing platform, called Google Earth Engine (GEE), to effectively address the challenges of big data analysis. In particular, this platform facilitates processing big geo data over large areas and monitoring the environment for long periods of time. Although this platform was launched in 2010 and has proved its high potential for different applications, it has not been fully investigated and utilized for RS applications until recent years. Therefore, this study aims to comprehensively explore different aspects of the GEE platform, including its datasets, functions, advantages/limitations, and various applications. For this purpose, 450 journal articles published in 150 journals between January 2010 and May 2020 were studied. It was observed that Landsat and Sentinel datasets were extensively utilized by GEE users. Moreover, supervised machine learning algorithms, such as Random Forest, were more widely applied to image classification tasks. GEE has also been employed in a broad range of applications, such as Land Cover/land Use classification, hydrology, urban planning, natural disaster, climate analyses, and image processing. It was generally observed that the number of GEE publications have significantly increased during the past few years, and it is expected that GEE will be utilized by more users from different fields to resolve their big data processing challenges.

A framework for extracting urban functional regions based on multiprototype word embeddings using points-of-interest data

Many studies are in an effort to explore urban spatial structure, and urban functional regions have become the subject of increasing attention among planners, engineers and public officials. Attempts have been made to identify urban functional regions using high spatial resolution (HSR) remote sensing images and extensive geo-data. However, the research scale and throughput have also been limited by the accessibility of HSR remote sensing data. Recently, big geo-data are becoming increasingly popular for urban studies since research is still accessible and objective with regard to the use of these data. This study aims to build a novel framework to provide an alternative solution for sensing urban spatial structure and discovering urban functional regions based on emerging geo-data – points of interest (POIs) data and an embedding learning method in the natural language processing (NLP) field. We started by constructing the intraurban functional corpus using a center-context pairs-based approach. A word embeddings representation model for training that corpus was used to extract multiprototype vectors in the second step, and the last step aggregated the functional parcels based on an introduced spatial clustering method, hierarchical density-based spatial clustering of applications with noise (HDBSCAN). The clustering results suggested that our proposed framework used in this study is capable of discovering the utilization of urban space with a reasonable level of accuracy. The limitation and potential improvement of the proposed framework are also discussed.

Quantifying the scale effect in geospatial big data using semi-variograms.

The scale effect is an important research topic in the field of geography. When aggregating individual-level data into areal units, encountering the scale problem is inevitable. This problem is more substantial when mining collective patterns from big geo-data due to the characteristics of extensive spatial data. Although multi-scale models were constructed to mitigate this issue, most studies still arbitrarily choose a single scale to extract spatial patterns. In this research, we introduce the nugget-sill ratio (NSR) derived from semi-variograms as an indicator to extract the optimal scale. We conducted two simulated experiments to demonstrate the feasibility of this method. Our results showed that the optimal scale is negatively correlated with spatial point density, but positively correlated with the degree of dispersion in a point pattern. We also applied the proposed method to a case study using Weibo check-in data from Beijing, Shanghai, Chengdu, and Wuhan. Our study provides a new perspective to measure the spatial heterogeneity of big geo-data and selects an optimal spatial scale for big data analytics.

A generalized supervised classification scheme to produce provincial wetland inventory maps: an application of Google Earth Engine for big geo data processing

ABSTRACTWetlands are important natural resources due to their numerous ecological services. Consequently, identifying their locations and extents is imperative. The stability, repeatability, cost-effectiveness, multi-scale coverage, and proper spatial resolution imagery of satellites provide a valuable opportunity for their use in various large-scale applications, such as provincial wetland mapping. To do so, it is required to (1) process and classify big geo data (i.e. a large amount of satellite datasets) in a time- and computationally-efficient approach and (2) collect a large amount of field samples. In this study, Google Earth Engine (GEE) and machine learning algorithms were utilized to process thousands of remote sensing images and produce provincial wetland inventory maps of the three Canadian provinces of Manitoba, Quebec, and Newfoundland and Labrador (NL). Additionally, using GEE, a generalized supervised classification method is proposed to produce a regional wetland map from a large area (e.g., a province) when lacking field samples. In fact, using the field data from only Manitoba and assuming that all wetlands in Canada have similar characteristics, the wetland maps were generated for the other two provinces. The overall classification accuracies for Manitoba, Quebec, and NL were 84%, 78%, and 82%, respectively, indicating the high potential of the proposed method for aiding provincial wetland inventory systems.

Revealing deep semantic commercial patterns: Insights from urban landscape depiction

Commercial functions play a significant role in economic growth and sustainable development. Influenced by socioeconomic changes which were investigated based on commercial behaviours using big geo-data, urban environment shows polycentric structures and further leads to more complex commercial landscapes. Thus, research considered environmental consequences of urban growth to better understand the commercialization process. However, only general commercial areas or certain commercial retail locations were investigated with the effects of urban environmental landscapes in existing studies. The decomposition of the commercial patterns such as accommodation and catering functions remains undiscussed. To fill this gap, this paper proposes a framework to decompose commercial patterns based on insight from different urban landscapes. First, we depict urban landscapes using spatial metrics, including the density of points of interest (POIs), percentage of landscape of buildings, landscape shape index of buildings and density of roads. Then, we examine the spatial distribution of the decomposed commercial patterns by delineating potential deep semantic topics. Finally, a structural equation model is utilised to determine the relations among urban landscapes and commercial patterns. The results show that four types of urban landscapes exhibit different spatial patterns, revealing various perspectives of commercial characteristics. Meanwhile, the decomposed commercial patterns, including those for catering and accommodation functions, display a heterogeneous distribution. These commercial patterns are directly affected by various configurations of buildings, POIs and roads. On this basis, suggestions are offered to improve commercial pattern development, including integrating urban landscape construction, organising the commercial pattern distribution and enhancing the harmony between the urban environment and commercial land uses.

Investigating Road-Constrained Spatial Distributions and Semantic Attractiveness for Area of Interest

An area of interest (AOI) refers to an urban area that attracts people’s attention within different urban functions through cities. The wide availability of big geo-data that are able to capture human activities and environmental socioeconomics enable a more nuanced identification of AOIs. Current research has proposed various approaches to delineate continuous AOI patterns using big geo-data. However, these approaches ignore the effects of urban structures such as road networks on reshaping AOIs, and fail to investigate the attractiveness and certain functions within AOIs. To fill this gap, this paper proposes a systematic framework to investigate the spatial distribution of road-constrained AOIs and analyze the semantic attractiveness. First, we propose an Epanechnikov-based kernel density estimation (KDE) with a bandwidth selection strategy to extract road-constrained AOIs. Then, we establish semantic attractiveness indices regarding AOIs based on the textual information and the number of review data. Finally, we investigate in detail the spatial distribution and semantic attractiveness of AOIs in Yuexiu, Guangzhou. The results show that road-constrained AOIs can not only effectively capture the human activity patterns influenced by urban structures, but also depict certain urban functions including entertainment, public, service, hotel, education, and food functions. This method provides a quantitative reference to monitor urban structures and human activities to support city planning.

URBAN GEO BIG DATA

Abstract. The paper deals with the general presentation of the Urban GEO BIG DATA, a collaborative acentric and distributed Free and Open Source (FOS) platform consisting of several components: local data nodes for data and related service Web deploy; a visualization node for data fruition; a catalog node for data discovery; a CityGML modeler; data-rich viewers based on virtual globes; an INSPIRE metadata management system enriched with quality indicators for each dataset.Three use cases in five Italian cities (Turin, Milan, Padua, Rome, and Naples) are examined: 1) urban mobility; 2) land cover and soil consumption at different resolutions; 3) displacement time series. Besides the case studies, the architecture of the system and its components will be presented.