Abstract
The aim of this study was the safety risk assessment of accidents that occur because of production dynamics in deep coal mining, one of the highly dangerous business class, with spatiotemporal GIS. In this study, accidents that occur at certain times during production in underground mining workplaces are discussed based on evaluations of spatial and temporal dimensions. Accordingly, various analyses have been handled using criminology science tactics and strategies such as 80–20, kernel density, space–time cube, and hot spot, which have been adapted to GIS tools. Implementations have been realized on 3D underground cartographic models covering accident data from deep coal mine workplaces in the Zonguldak-Kozlu basin in Turkey between 2019 and 2021, at an average elevation of – 400 m. Space–time cube, kernel density, and time series analyses suggest that the Annual Bonus Incentives, which are repeated every year and given before the new year, increase the psychosocial risks for employees. It has been determined that this situation can lead to problems such as poor concentration among miners and loss of focus in some shifts. Additionally, significant relationships were found between the number of accidents and monthly production progress rates as a temporal parameter. The results demonstrate that applying a reactive approach with spatiotemporal GIS to accidents in deep coal mines is becoming increasingly important.
Similar content being viewed by others
Avoid common mistakes on your manuscript.
1 Intruduction
Today, coal is still the most widely used fuel type in energy production and industry. Mines, and especially underground mines, contain so many hazardous environments such as flammable, explosive, dusty, hot, wet, humid, and poorly lit. In addition, those working in these closed and narrow environments are at risk of collision, jamming, and material falling on them. Out of the other reasons of coal mine accidents are lack of risk evaluation, production pressure, not taking lessons from previous accidents, insufficiency in supervision (internal audit) services in mining businesses, ineffective supervision of public agencies, lack of vocational training, workplace safety culture, and psychosocial hazards [1]. Researches show that the majority of accidents occur in coal mines due to the reasons stated, and unfortunately injuries and deaths in mining workplaces, especially in developing countries, still take place at an alarming rate [2,3,4,5,6,7,8].
Every year, 350,000 workers in the world and 1200 workers in Turkey lose their lives because of occupational accidents. The ILO also estimates that 395 million workers worldwide sustained non-fatal work injuries [1, 9]. The numbers of workers’ deaths per 100,000 tonnes coal production in occupational accidents in Zonguldak Hard Coal Field are 2.83 in private and illegal pits and 2.03 in Turkish Hardcoal Authority [10]. Mining is one of the fields where deaths related to occupational accidents are most commonly seen in Turkey.
Underground mining workplaces are in the category of very dangerous workplaces according to the European Union regulations. European Social Policies, the judgment C-84/94 of 12 November 1996 of The European Court of Justice (ECJ), and the World Health Organization (WHO) gave the following definition for health. Health is the state of complete physical, mental and social well-being and not merely the absence of disease or infirmity. This definition emphasizes the psychosocial state of employees. Therefore, it is important to carry out risk analyses in terms of occupational health and safety and to take actions quickly. These analyses also include visual elements [3, 4, 11]. In addition, mining workplaces are always under high risk not only in terms of human and environmental factors but also the accidents that may occur as a result of these risks. These can cause serious physical and mental damage and even lead to death [12, 13].
ISO, the International Organization for Standardization based in central Geneva, has published a guide at an international level titled ISO 45003 “Psychological Health and Safety in the Workplace” in 2021, which provides guidance on psychosocial hazards and risks in the workplace under the ISO 45001 Occupational Health and Safety Management Systems Standard [14].
Psychosocial hazards are factors that cause more stress than the employee can cope with. According to ISO 45003, psychosocial risk is the likelihood of exposure to psychosocial hazards related to work and the severity of injury and health impairment that these hazards can cause (ISO 45003, p. 3, Article 1). A study conducted on 252 miners in Pakistan identified high job demands and pressure in the workplace as sources of psychosocial risks affecting musculoskeletal disorders [15].
In the assessment of psychosocial risks in the workplace, incidents are approached with reactive and proactive methods. The proactive approach aims to take precautions by anticipating accidents, dangers, and risks, and to eliminate or minimize the damage that may occur as a result [16,17,18]. But unfortunately, in risky applications, such as underground coal mining, there are often unpredictable hazards and events. It is not practical to pre-plan for every possible contingency underground, and therefore making proactive assessments challenging [18].
The aim of the reactive approach is generally based on the examination of the situation after the near-miss or accident and to take various measures as a result of this examination [3, 16,17,18]. As a result of post-event evaluations, harm to the employee, and loss of time, work, and equipment cannot be prevented in the first step. Hence, it is important to adopt a reactive approach to develop protective strategies and precautions for unexpected events or to prevent future incidents or accidents.
Managers in mine workplaces can use spatiotemporal GIS techniques to conduct reactive risk assessments based on criminology tactics and strategy. This allows for immediate intervention in disruptions to production planning, rapid responses to unexpected events, and urgent updates to risk management plans. By re-examining production planning, solutions to the problems can be found [19, 20].
The abovementioned situations require that accidents occurring in underground mine workplaces be examined with a reactive approach and their temporal dimension. Spatiotemporal GIS provides a powerful support mechanism for decision makers, especially in the public sector, in terms of safety [21,22,23,24,25,26]. Thus, smart technologies increase mining efficiency in underground coal mines and reduce workload and accidents [27,28,29]. By using space and time analysis and 3D visualization technology, the spatiotemporal model of mine accidents is strategically made understandable. Applications primarily require time-based modeling, analysis, and visualization of the geometric and/or semantic changes of geographical phenomena. For the temporal pattern, this situation should be evaluated in such a way for receiving answers in all its dimensions for such questions as “where and when did the change occur,” “what kind of change occurred,” “what is the rate of change,” “what was the changeover period,” “what is the duration of change,” “did the change occur in different places at the same time or at different times in the same place,” “why did it happen,” and “how did it happen” [21].
There are many important studies carried out with spatiotemporal GIS in different fields, which will serve as examples for the investigation of mining accidents. A list of some of the important ones is given in Table 1.
When Table 1 is examined, it is seen that the study contents are directed to four basic data groups that constitute the input to the Temporal GIS [46]. These are as follows:
-
Data on motion and moving objects, such as storm motion,
-
Discrete events such as trafic accidents,
-
Observations based on station or sensor records installed on them, such as a river,
-
Data on change, such as a forest fire
All given temporal GIS structures are applicable for mine transportation, mine accidents, mine ventilation, and mine fire. This study utilized the discrete events structure for mine accidents. Unfortunately, spatial and temporal analysis and GIS mapping techniques are not used for a reactive risk assessment regarding accidents in mine workplaces today. In this case, decision-making managers try to reach a conclusion by evaluating each event independently or making assumptions that do not go beyond predicting the relationship between events.
As with crime analysis, identifying the regions where accidents occur most frequently on maps and supporting these maps with some graphics and visual models can ensure that risk assessments always have a high success rate. This situation will increase efficiency by ensuring that employees and authorized personnel are assigned the correct number of tasks at the right time and place, and so accidents will be prevented. For this study, accidents occuring due to production dynamics in deep coal mining, one of the highly work class, are handled for the purpose of reactive risk assessment in the form of discrete events and data on spatial change; and therefore, a database has been created. The results are evaluated from the perspective of tactics and strategies adapted from criminological science.
2 Materials and Methods
2.1 Spatial and Temporal Data Infrastructure
Spatial-temporal analysis and modelling, it is generally considered to be a synthesis of GIS that defines space and models linked to dynamic processes in the natural environment [47,48,49,50]. In order to create spatial models, it is necessary to obtain data about the environment in which the processes ocur. The use of spatial models in temporal analyses allows the intelligibility of such analyses and the rapid reaction of the directors at the decision maker level to the assessment of risks.
In mining workplaces, a Mining Information System (MIS) can be used for spatial data infrastructure, which defines the underground environment well. In this study, an MIS was utilized, which was created with the microstation software infrastructure for the Turkish Hard Coal Enterprise operating in Zonguldak-Kozlu and is active. The subject was examined for the Kozlu region in the Zonguldak hard coal basin of Turkey, which is shown in Fig. 1. The analysis focused on four underground production panels operating at deep elevations between 2020 and 2022 in the region. Studies on production panels where accidents are analyzed, between depths average of – 300 m/ − 460 m sea level, coal seam thicknesses varying between 2.30 and 8.30 m, and an average inclination angle of 32°. The average daily production progress in the panels was 0.2 m. Space–time analyses were made by evaluating 227 accident data that occurred during production in different shifts together with 3D cartographic models derived from mine production maps (MPM).
The spatial data presented in Fig. 2 was obtained from MPMs and accident reports. Since MPMs are legal graphic documents reflecting the status of the mining enterprise from the past to the present, it is difficult to turn them into any other documents or change them. All current applications should be applied on these original graphic MPMs. Mining has been carried out in Zonguldak since 1895 and thousands of MPM have been produced. Underground topographical measurement data are first drawn on these graphic documents. Subsequently, the document is digitally scanned to obtain a raster image and vector data topology is created by raster to vector conversion (Fig. 3). Vector models are used for their 3D cartographic representations. This implimentation allows the MPMs to be used in the Mine Information System. In addition, documents of accident reports are also digitally scanned, and drawings and attribute data describing the accident and the scene are associated with the 3D cartographic models.
Temporal data presented in Fig. 2 can be classified into two types of functional time: processing time (recording time) and valid time (accident time, event time). Transaction time indicates the date when the event is recorded in the database, while valid time indicates the time period during which the event exists. In temporal GIS applications that analyze discrete events, it is necessary to have both record types in the database, depending on the type of query, to avoid confusion. Events that involve both types of time are referred to as double time. Additionally, it is possible to refer to the time of measurement. When the exact time of an accident is unknown, it is estimated based on the time of field measurements taken after the accident [50].
For the accident analysis with spatiotemporal GIS, the MPM of the relevant production panels taken from MIS is given in Fig. 3. These vector models were converted from scanned images of the MPMs given in Fig. 4. In this cartographic model, accidents in the production panels are shown by placing them in their real positions. In the second stage of the study, there are prepared 3D cartographic models of the mine panel produced by the longwall method on the coal seam called “Acilik” in Fig. 5. 3D cartographic model of the panel that produces mines using the slicing longwall method on the extra thick coal seam called “Çay” is shown in Fig. 6. In the vector models of the panels, the production date is written on the monthly production area and displayed as parcel sectors. The width value of each parcel on the map, measured from the vector map, gives a monthly production velocity (meters/month). In addition, the depths of production carried out below sea level are recorded on the cartographic models. With these height values and the panel width values measured from the cartographic model, the slope of the parcel, which is the monthly production area on the panel, and the slope angle studied are determined.
The first step for the accident data was to group the attribute data in the accident reports and transfer them to the database. In terms of ethics, the personal information of the employees was not used in the project, and their ID numbers were determined by tagging them to be used instead. In this way, the attribute data in each row has been made special and distinguishable for each event. Event date has been recorded in the database in accordance with ISO 8601:2004 standards [47, 51, 52], in “DD/MM/YYYY” format and event time in “hh:mm” format. The explanation of the data in the created attribute table is given in Table 2.
2.2 Methods Used for Spatiotemporal Analysis
The process of converting spatial and temporal data into information using statistical methods is referred to as analysis in GIS applications [53,54,55]. Unlike querying from a database, new information is obtained as a result of analysis. Injury accidents that occurred during deep underground mining practices were examined in this study. Spatial and temporal data were obtained by using vector and temporal data regarding 3D cartographic models of production areas and accident reports. These data were evaluated with tactics and strategies of spatial and temporal GIS developed from criminological science. The use of criminology in spatial and temporal GIS stems from the overlap between crime analyses and analyses of events such as accidents, diseases, natural events, and fires that occur in a specific location. The mining accidents dealt with in this study are evaluated against the act of crime in the criminology.
Crime analysis in criminology is typically divided into two categories: tactical and strategic [56]. Tactical use involves analyzing crime data to prevent future incidents. The evaluation and analysis produced for tactical use aim to answer the questions of how, where, and when. The information produced for strategic use, based on the analyses, can be used by senior managers to make decisions and identify and prevent crime tendencies and types. When analyzing a crime, three basic applications are prominent [57].
-
1.
Proactive approaches to early offense prevention.
-
2.
Reactive approaches to elucidating existing offenses: Involve profiling crimes and criminals, including studying the methods used to commit crimes, the regions in which they occur, and the time periods in which they are committed. To clarify mine accidents, it is necessary to profile the accidents and victims, and analyze the time periods and practices involved.
-
3.
Determination of crime tendencies: Economic, social, and cultural changes in society can lead to changes in the types of crimes committed over the long term. Determining the trends in mining accidents, including the type of accidents that occur during each shift and time period, will be effective in preventing future accidents.
In criminology, it is important to analyze and classify (by classification here is meant not the classification of the type of crime, but the method, time, or manner in which it was committed) events related to crimes. Failure to do so may lead decision-makers to assume that all events are independent of each other or to make unfounded assumptions [57]. Assessments made with crime analysis will always have a high success rate. Therefore, efficiency and productivity will increase by assigning the appropriate number of employees and authorized personnel at the right time and place to prevent crimes. The same applies to mining workplaces. The crime analysis phase involves the following steps.
-
1.
Data collection methods may include mathematical data, observations, and questionnaires.
-
2.
Data classification is a crucial stage in preparing data for analysis in a GIS structure. GIS applications enable relationship analyses between individuals, crime mapping, and statistical analyses using crime-related data.
-
3.
Analysis involves evaluating classified data within the framework of the questions to be answered and associating complex information structures that may appear independent. The methods used for analysis vary depending on the nature of the question or the intended audience. Various methods can be employed in crime analysis, taking into consideration the intended audience. For instance, maps can be used to identify regions with the highest crime rates, and these maps can be accompanied by graphics and displayed in public areas for the benefit of the community. For this reason, Gottlieb (1998) and Gürer (2004) highlight two main issues in data analysis [57, 58]. Firstly, they draw attention of analyses that use mapping techniques to make geographical associations and regional evaluations. Secondly, they address the frequency of similar events, the determination of the similarities of the factors in the events, and the techniques that allow for different queries to be made from the available data.
-
4.
Announcement of offenses to the relevant parties
-
5.
Feedback and evaluation
In this study, the tactics and strategies used for steps of the analysis of injury accidents and developed for crime analyses in crimonology are given in Fig. 7, and subsections presented more details.
2.2.1 Kernel Density Tactical and Strategic Analysis
A probability density function is used for kernel density spatiotemporal analysis. It is the process of determining the density of points falling within a circle with a defined radius and the point density changing as it moves away from this source. With this analysis, useful results can be obtained in determining the places where accidents in underground mine workplaces are hot spots or where they occur frequently. A kernel function is used to properly fit a conical surface to each accident point vector data. Where Xi is n number of independent identically distributed population samples, the kernel estimator of the probability density function can be calculated from Eq. 1 [59].
In this equation, K is the kernel function and h is the window width, smoothing parameter, or bandwidth. The kernel function determines the weights of the estimation together with the bandwidth. Usually, K is a probability density function. Barrier affect can be used to determine a population-based core density in underground mining areas. For example, in a longwall production area, gate roads may or may not be taken into account. If not, a pixel-based density zone map is created by calculating a density value on the production area, whereas when you take gate roads into consideration, only the population density within the road (for example, sensor density) is calculated and mapped (Fig. 8).
2.2.2 Hot Spot Tactical and Strategic Analysis
Hotpoint spatiotemporal analysis is used to determine the intensity and distribution of events occurring in an area over a specific time period. This analysis method uses statistical methods to test whether events are random, which regions are more exposed to events, and whether events are spatially correlated with each other. Given a set of weighted features, Getis-Ord identifies statistically significant hot spots and cold spots for the spatial distribution of accidents with the help of the Gi* statistic. Equations 2, 3, and 4 provide the formula for the Getis-Ord statistic [40, 60].
In these equations, xj represents the attribute value for object j; wi, represents the spatial weight matrix between objects i and j, if the distance from a neighbor j to the feature i is within the distance, wij = 1; otherwise wij = 0; n represents the number of features. In pattern analyses such as hot spots, it is necessary to first define a null hypothesis. The pattern analysis assumes complete spatial randomness as the null hypothesis, for either the features themselves or the values associated with those features. The z scores and p values from the pattern analysis tools indicate whether you can reject the null hypothesis. The p value represents the probability that the observed spatial pattern was created by a random process when using pattern analysis tools. A small p value indicates that it is highly unlikely that the observed spatial pattern is the result of random processes, allowing for the rejection of the null hypothesis (Fig. 9).
For instance, if the p value is low and the z score is high for the relationship between location and the number of injuries during the analysis of the pattern of injuries in accidents that occur in a hazardous mine production panel, it is assumed that the injuries are not random and that there is a problem originating from the location [61].
As can be seen from Fig. 9, the critical z score values for a 95% confidence level are − 1.96 and + 1.96. The uncorrected p value associated with a 95% confidence level is 0.05 (0.95 = 1 − p). For example, if the z score of accidents occurring in an underground mine production panel is 0.7 and falls between − 1.96 and + 1.96, the uncorrected p value will be greater than 0.05. Therefore, the null hypothesis cannot be rejected for the pattern of accidents observed within the panel because the pattern observed could be the outcome of random spatial processes.
The z score (in standardized normal distribution is the standard deviations value) returned by the Gi statistic for each feature in the dataset indicates the intensity of clustering of high or low values. High values (hot spots) and low values (cold spots) define statistically significant spatial clusters, and the z score and p value and confidence level are generated for each feature in the input feature class [62, 63]. The z scores and p values indicate the spatial clustering of features with high or low values. Hot spot analyses work by examining each feature within the context of its neighboring features. A feature with a high value may not necessarily be a statistically significant hot spot. For a feature to be considered a statistically significant hot spot, it must have a high value and be surrounded by other features with high values. The sum of a feature and its neighbors is compared proportionally to the sum of all features. If the local sum is significantly different from the expected local sum and this difference cannot be attributed to random chance, a statistically significant z score is obtained.
The results of Gi* statistics at different confidence levels provide outputs on which hot spot is more effective than the others [64]. When interpreting the results of the hot spot analysis at different confidence levels, a hot spot at a 99% confidence level means that it is a more risky place for events than a hot spot at 95% and 90% confidence levels. In this method, the application is implemented by selecting appropriate parameters in the point layer containing the locations of the events. The search window size, weighting factor, and the selected statistical test are considered parameters. The output map shows results colored with warm and cold colors according to the intensity of the events. In the output map display, points are specified as hot points, cold points, or neutral points (Fig. 10). Hot spots are regions where events are concentrated according to the application structure and are spatially correlated and do not occur randomly. (However, it cannot be said here whether the event is temporally random or not. The normal distribution structure should be examined with time series.) Cold points are areas where events are few or infrequent. Neutral points are regions where events are randomly distributed and statistically significant results do not occur.
2.2.3 Buffer Point Density Tactical and Strategic Analysis
In buffer point density analysis, a spatial analysis is performed based on examining the distances of an event to other surrounding events [65]. In the process, a new buffer zone with a polygon feature is created at an impact radius (for instance, when considering the width of the excavation area in a mine production panel) around an event that is considered as a reference. This region is mapped as an occlidian buffer (on a 2D plane) or a geodesic buffer (on an ellipsoid, sphere, or geoid). In the underground mining production area, the geographical detail or event in the point vector data structure is accepted as the center and a circle of the desired radius is created.
The composite structures of the buffers formed when all the buffers within a certain radius of action in the circle type related to the point data of each event are resolved are evaluated as regions where events are concentrated and not random. In Fig. 11, the buffer point density for point data type events on a longwall type mining production panel, for example, points where the dust density is high on the panel, is shown as an example.
2.2.4 80–20 Tactical and Strategic Analysis
The 80–20 (Crime Analysis and Safety) spatiotemporal analysis method was developed for GIS applications, benefiting from the science of criminology. The 80–20 rule, also known as the Pareto principle, is an inference that claims that 80% of the consequences of events arise from 20% of all possible causes for any event. In practice, cluster positions are determined by creating a graduated symbolological layer based on the spatial locations of the input features and the number of events. As a result of this analysis, a cumulative percentage area is calculated to identify event locations that occur disproportionately. The 80–20 rule is based on the principle that the vast majority of events occur in a small minority position. It is a theoretical concept that, according to criminology, 80% of events (e.g., crimes) occur in 20% of locations (e.g., crime scenes). GIS applications support this concept [63, 66].
2.2.5 Tactical and Strategic Analysis of the Number of Accident Incidents
The number of accident incidents method is the counting of events that occur in a certain place and time by enclosing them in a polygon. To do this, point data representing events are collected in polygons to create a new layer. Since a polygon (parcels) is the geometric shape of a longwall mine production panel at the end of a month on the mine production map (MPM), in this application, parcels on 3D cartographic models showing monthly production in underground hard coal production areas were used. Monthly progress rates may vary depending on production conditions. For this reason, production parcels in the form of polygons of different sizes are formed on the map according to the monthly production rate (Fig. 12). The number of events and event types occurring within these polygons can be separated. If focused on a specific type of event in the production areas of mining enterprises, this method can produce useful results in terms of risk management and safety.
2.2.6 Time Series and Space–Time Cube Tactical and Strategic Analyses
Time series analyses focus on purely temporal analysis of localized data. Geostatistics and spatial statistics deal mostly with the spatial description of data. These methods are extensions of time series methods in the spatial domain [66, 67]. Widely used conventional static GIS neglects the fundamental dynamics of natural processes over time. It does not account for important cross-correlations and causal dependencies in the combined space/time domain. Therefore, it significantly limits the practitioner’s understanding of the situation of interest. This situation reduces the predictability capabilities of GIS. In the common absolute conceptualization, for a three-dimensional ZGIS, two dimensions constitute the Euclidean space and one dimension constitutes the time orthogonal. In space and time, the concept of movement (or inactivity) is responsible for relating space and time.
Space and time help reveal the locations, situations, and histories of objects for meaningful explanations. It also helps develop models to make predictions through space and time relationships, empirical regressions, computational simulations, or machine learning algorithms. With time-sensitive layers, changes and trends of measurements over time can be determined. Thanks to temporal GIS models created with temporal analysis methods, the flow of events over time can be monitored and examined in detail, allowing temporal inferences to be made. The behavior of time series data for observations measured regularly over time is investigated with trend analysis methods. Trend analysis is performed with parametric t-test and non-parametric Mann–Kendall test. The t-test is a parametric statistical trend test that assumes the variable has a normal distribution. The Pearson correlation coefficient between observations and time is determined and calculated to test the null hypothesis according to the t-distribution. The Mann–Kendall test, on the other hand, makes no assumptions about the distribution of the variable. The Mann–Kendall test is a non-parametric test based on ranks used to evaluate the presence and significance of a trend, and is frequently used to detect trends in time series data [68]. Mann–Kendall statistics are used in temporal GIS. The Mann–Kendall test is based on the null hypothesis that the sampled data are independent and uniformly distributed. This means that there is no correlation in the trend between data points. The alternative hypothesis is that a trend exists in the data. Mann–Kendall statistics are also used for analyses performed by creating a space–time cube. The first step in the Mann–Kendall method is to calculate the variable S, which is the sum of the differences between the data points shown in Eq. 5.
In the equation, n is the number of points in the data group, Xi is the data value in time series i, and Xj is the data value in time series j. The sign of the (Xi – Xj) value is determined as in Eq. 6. When n ≥ 8, the S statistic is approximately normally distributed with mean and variance (corrected for ties) and is determined as in Eq. 7.
In Eq. 7, ti is the number of ties in time series i; n, number of data points; m, number of connected groups. A linked group is a set of sample data with the same value. In cases where the sample size is n > 10, the standard normal test statistic is calculated using the equation Zs. The normally distributed S statistic allows the calculation of the standardized test statistic value (Zs) and the Mann–Kendall test, and the Zs statistics are calculated as shown in Eq. 8.
The standardized Zs statistic has a normal distribution, with a mean of 0 and a variance of 1. A positive or negative value of Zs indicates increasing or decreasing trends respectively [29, 30]. The significance of the trend is determined by the z-test (see Fig. 8). For example, if the z score is greater than + 1.96 or less than − 1.96 at a 95% confidence level and the p value is less than 0.5, the trend is considered significant. Space–time cube and time series analysis applications are used as one of the most effective methods for dynamic visualization. The base of a space cube represents the location in space, and the height represents time. The space–time cube consists of trajectories and location time prisms that show movement between locations within a time period. The basis of space–time cube and temporal analysis is based on Mann–Kendall statistics. A series of points are collected in space time bins and summarized in a data structure. In this way, cubes of point input features are formed, with x and y dimensions representing space and t dimension representing time. A summary area statistic is calculated by counting the points in each compartment of the cube, and the trend of the time interval values at each location is measured using the Mann–Kendall statistics and the time series graph is plotted. In Fig. 13, the space–time cube model is given on the underground mine production map.
Space–time cubes that are displayed on MPMs summarize events that occur in specific time periods in mine production panels. Events that occurred at different time are identified by their X,Y horizontal Euclidean coordinates and time dimension, and have been represented by red dots in Fig. 13. Mann–Kendall statistics are used to color the events on space–time cubes. If events occur in different locations during the same time period, they are summarized in red on the horizontal cubes. If events occur in the same location during different time periods, they are summarized in red on the vertical cubes.
3 Results and Discussion
In the spatiotemporal analysis, the distribution of injury levels was first examined, and the findings are given in Fig. 14 as a thematic map. According to this analysis, 72 of the accidents involved “calf and foot injuries” known as K6. The incidence of this type of accident in the total number of accidents is 33%. After K6, the highest injury levels are followed by K5 (hand and finger injuries) with 31%, K7 (head injuries) with 13%, K4 (shoulder and upper arm injuries) with 13%, K1 (neck and spine injuries) with 7%, K3 (hip and knee cap injuries) with 2%, and K2 (rib cage injuries) with 1%. It has been determined that 64% of a total of 227 accidents seen in Fig. 14 occur in injury types called K5 and K6. It is thought that this situation is caused by the equipment they work with and that they are inexperienced and/or have low education level in the use of this equipment. It may also be an indicator of carelessness in their use due to psychosocial problems.
In total, 225 of the injured employees are primary, secondary and high school graduate workers, and two of them are university graduate technical personnel. The distribution analysis results of the injured according to the years of experience are given in Fig. 15. Based on this analysis, 74 workers, which accounts for 33% of all injured workers, have been working in this occupation for a decade.
The second place is seen as those who have been working for 11 years with 40 employees, and those who have been employed for 1 year with 39 employees in the third place. It cannot be said that the high rate of accidents among employees with 10 and 11 years of experience is due to their lack of adequate training. On the contrary, it is thought that this is an indication that there are some psychosocial problems in employees with 10 and 11 years of experience. It can be said that the situation of 32 new recruits and 39 people with 1 year of experience stems from the lack of training and experience regarding the working environment and method.
Production in the basin is carried out as three shifts. The first shift is a production shift, between 00:00 and 08:00. The second shift is a repair and preparation shift, between 08:00 and 16:00. The third shift is also a production shift, and it operates between 16:00 and 00:00. The spatiotemporal distribution of the accidents in these shifts is given in Fig. 16. In total, 37% of the accidents occurred during shift 2. Shift 3 is in the second place with 33%, and shift 1 is in the third place with 30%. The graphs of the accident numbers according to the hours in the working period of each shift and kernel density maps of accidents in shifts are presented in Fig. 17.
Considering the distribution of accidents according to hours, it is noticed that an increase in the number of accidents was detected in the first hours of the shift in shifts 1 and 3. In shift 2, the accident increase was determined at noon in the middle of the working period, and when the density analysis results are evaluated, it is understood that the accidents intensified in the Çay coal seam 2020 panel in the fifth slice of slicing longwall production. This situation may stem from reasons such as not being able to adapt to the place and work, loss of concentration, and lack of attention as employees start work in the evening and night hours in shift 1 and shift 3, and coinciding with an hour close to lunch in shift 2.
As another application of spatial and temporal analysis, the production rates in underground coal production areas with a dynamic structure were evaluated. The monthly production rates of 2019 and 2020 production panels in the two coal seams, which were determined by cartographic models in the study, are given in Table 3. In addition, a thematic vector map was produced as a result of the application made with the number of events analysis (Fig. 18). The increase in the number of accidents occurring in monthly production progress is shown in dark red on this map.
When the map in Fig. 18 is examined, it is seen that the number of accidents increases in the months when the production rate increases (24.1, 16.6, 11.8, and 10.8 m/month) and when the monthly production rate is as low as 2 m/month. In terms of occupational safety, this implies that the number of accidents decreases in the months when production is carried out at normal speeds. It is thought that being in a struggle for reaching the planned speed in the months when the production decreases or the desire to get an extra premium at the end of the year when the production is intense dominates this situation and constitutes a psychosocial problem.
According to 80–20, a GIS analysis technique developed on a basic theory of criminology science, it is foreseen that 80% of crimes occur in 20% of the space. This spatial and temporal GIS analysis technique has also been used in the assessment of accidents in underground deep coal mines and the assessment results are presented in Fig. 19. Findings verify this theory. It is seen that nearly 80% (181 accidents) of the accidents occurred in the 2019 and 2020 panels of the Çay coal seam, where the slicing longwall method is applied, which constitutes about 20% of the space.
Optimized hot spot and buffer analyses were also performed to confirm this finding (Figs. 20, 21, and 22). Also, findings reveal the dangers of production of the fifth floor slice of the Çay coal seam, which is very risky and difficult to apply for the years 2019 and 2020, and the adaption difficulties of the employees to the production place and method.
Statistical results for the hot spot analysis of accidents are presented graphically in Fig. 21. In hotspot analysis, a low negative z score and a small p-value indicate spatial clustering of low values. The higher (or lower) the z score, the more intense the clustering. A z score close to 0 indicates no significant spatial clustering [61, 69]. The graph in Fig. 21 shows that in the production panels where accidents are examined, both low and high values exhibit spatial clustering. The null hypothesis is rejected and it is observed that accidents are clustered in a significant (unfavorable) spatial structures in production panels. Accidents the mine roadways in rock, it was determined that the p values were greater than 0.1 and the null hypothesis was accepted and these accidents were random.
3D space–time cube analyses on the temporal development of occupational accidents are possible. In this application, in order to evaluate occupational accidents, the occurrence dates of 227 accident data were taken into account, and space–time cubes were created for the accidents that occurred in a 2-year time period. While creating the space–time cube, monthly progress was taken into consideration as time steps and spatial–temporal density analyses of the accidents were made with the cubes created.
In Fig. 23, the cubes obtained according to the space–time analysis of the accidents are shown on the raster cartographic models. A summary field statistic was calculated by counting the points in each division of the space–time cubes in this application, and the trend of the time interval values at each location was measured using the Mann–Kendall statistic (red cubes obtained on the bars in Fig. 23).
When time series of the space–time cube analyses in Fig. 23 were examined, it was determined that the z scores of the normal distribution were between – 1.96 and 1.96 at the 95% statistical confidence level and the p value (probability) was < 0.05 (Fig. 24). These statistical values show that accidents occur randomly and individually, and are not a mass accident. However, when the time series graph in Fig. 23 is examined, it has been determined that production stopped during the COVID-19 pandemic period between March 30, 2020, and June 1, 2020; and therefore, there were no work accidents.
In the space–time cube analysis, accident densities that occurred in the same time period but in different places have been determined. In each production panel, it is seen that the accidents are concentrated in the time period close to the new year. Cubes, shortly before Christmas 2019, 2020, and 2021, turn red and create a hot spot. When the reason for the concentration of accidents in the time period coinciding with the end of the year is investigated, it is noticed that the company made an extra payment to its employees in exchange for targeted gross production on December 27, apart from the salary, called Annual Bonus Incentives. This creates a positive effect for employees as well as psychosocial pressure which leads to the risks.
4 Conclusion
Upon examining the findings of the hot spot analysis and kernel density analysis, it was determined that the sliced longwall production areas on the Çay coal seam create hot spots and density. In these analyses, low p values and high z scores indicate a higher incidence of accidents in these areas. This phenomenon has also been identified in the 80–20 analysis map, buffer point density analysis map, and accident frequency analysis maps. In particular, it is necessary to increase the level of in-service training of the personnel working in this field. According to the Mann–Kendall statistic, space–time and time series analysis was conducted for accidents that frequently occur in certain regions, and statistically significant hotspots (red cubes) were observed in the cubes during the time period corresponding to the new year. It is seen that the Annual Bonus Incentives, which are repeated every year and given before New Year’s, create a high demand for work and put pressure on work speed, so psychosocial risks for employees are increasing. It has been identified that as a result of this problems such as inability to concentrate sufficiently and loss of concentration arise during shifts at certain time intervals by kernel density and shift analyses.
It is observed that in shift 2, known as the Repair and Preparation shift between 08:00 and 16:00, work accidents intensify before noon, while in other shifts, the intensity is at the start of the shifts. It is thought that it will be beneficial for the prevention of accidents by warning the employees to be more careful before the lunch breaks in shift 2, and during the starting times of the other shifts. In addition, these issues should be particularly emphasized in the training provided within the scope of occupational health and safety services.
Similar results were achieved within the scope of the “Capacity Building of Occupational Physicians for the Assessment and Prevention of Psychosocial Risks in Mining Sector” project [70], which started on February 1, 2021, in the Zonguldak Hard Coal Basin with the financial support of the European Union and the Republic of Türkiye, aiming to contribute to the psychosocial safety of mine workers (Project number: EuropeAid/162207/ID/ACT/TR, Contract number: TREESP1.1OHSMS/P-03/13, Contract value: EUR 188686). In this examination conducted with occupational safety specialists and employees, it was determined that employees have to cope with stress because of the tasks they need to complete in order to receive their salaries and bonuses, which creates psychosocial risks [71].
According to the results obtained from temporal and spatial analyses, accidents may vary depending on the time and the working environment in underground workplaces. For safety reasons, it is necessary to make the spatiotemporal GIS applications continuous in these workplaces and to present the results obtained as a report to the decision-makers by making new analyses according to the changing working area and conditions.
When the results of the Incident Number Analysis application and the monthly production progress rates were compared, the relations between the monthly production progress on the panels and the intensity of the accidents were found. It is seen that there is an increase in the number of accidents at low and high speeds and a decrease in the number of accidents at average speed levels. For this reason, it is recommended that the monthly panel progress determined in the implementation project be carried out according to the procedure without changing the production rates.
With this study, it is also aimed to contribute to risk management and assessment studies with 3D cartographical mine model and spatiotemporal analyses to accidents. When the application is evaluated in terms of national and international mining safety legislations within the scope of occupational health and safety, it has been seen that the application of this kind of approach in accidents in deep coal mines is important. It is thought that with this approach, measures can be taken for unexpected events or for the prevention of future events or accidents, protective strategies can be developed, production can be replanned, and decision-makers can respond quickly to events.
Data Availability
Data analyzed during the current study are available from the corresponding author upon reasonable request.
References
Çağla F, Küçük U, Ilgaz A (2015) Causes of coal mine accidents in the world and Turkey. Turk Thorac J 16(1):9–14. https://doi.org/10.5152/ttd.2015.003
Sari M, Duzgun HSB, Karpuz C, Selcuk AS (2004) Accident analysis of two Turkish underground coal mines. Safety Sci 42(8):675–690. https://doi.org/10.1016/j.ssci.2003.11.002
Maiti J, Khanzode VV (2009) Development of a relative risk model for roof and side fall fatal accidents in underground coal mines in India. Safety Sci 47(8):1068–1076. https://doi.org/10.1016/j.ssci.2008.12.003
Khanzode VV, Maiti J, Ray PK (2011) A methodology for evaluation and monitoring of recurring hazards in underground coal mining. Safety Sci 49:1172–1179. https://doi.org/10.1016/j.ssci.2011.03.009
Amponsah-Tawiah K, Ntow MAO, Mensah J (2016) Occupational health and safety management and turnover intention in the Ghanaian mining sector. Safety Health Work 7(1):12–17. https://doi.org/10.1016/j.shaw.2015.08.002
Roghanchi P, Kocsis KC (2018) Challenges in selecting an appropriate heat stress index to protect workers in hot and humid underground mines. Safety Health Work 9(1):10–16. https://doi.org/10.1016/j.shaw.2017.04.002
Tetzlaff E, Goggins KA, Pegoraro AL, Dorman SC, Pakalnis V, Eger TR (2020) Safety culture: a retrospective analysis of occupational health and safety mining reports. Safety Health Work 12(2):201–208. https://doi.org/10.1016/j.shaw.2020.12.001
Chong HT, Collie A (2022) The characteristics of accepted work-related injuries and diseases claims in the Australian coal mining industry. Safety Health Work 13(2):135–140. https://doi.org/10.1016/j.shaw.2021.12.701
ILO-Ankara (2023) ILO home>Regions and countries>ILO Office for Türkiye>News, Nearly 3 million people die of work-related accidents and diseases. Available via DIALOG. https://www.ilo.org/ankara/news/WCMS_902220/lang--en/index.htm
Sarıkaya İ (2013) Murder, not accident: workers’ deaths in the subcontracted pits of the Zonguldak Coalfield. Eğitim Bilim Toplum Dergisi 11:81–100
Ajith MM, Ghosh AK, Jansz J (2020) Risk factors for the number of sustained injuries in artisanal and small-scale mining operation. Safety Health Work 11(1):50–60. https://doi.org/10.1016/j.shaw.2020.01.001
Erdoğan HH (2016) A quantitative risk assessment methodology for occupational accidents in underground coal mines: a case of Turkish Hard Coal Enterprises. Dissertation, Middle East Technical University
Jha A, Verburg A, Tukkaraja P (2022) Internet of Things based command center to improve emergency response in underground mines. Safety Health Work 13(1):40–50. https://doi.org/10.1016/j.shaw.2021.10.003
ISO 45003 (2021) OHomepage>BS Standards>13 ENVIRONMENT. HEALTH PROTECTION. SAFETY>13.100 Occupational safety. Industrial hygiene>BS ISO 45003:2021 Occupational health and safety management. Psychological health and safety at work. Guidelines for managing psychosocial risks. Available via DIALOG. https://www.en-standard.eu/bs-iso-45003-2021-occupational-health-and-safety-management-psychological-health-and-safety-at-work-guidelines-for-managing-psychosocial-risks/?gad_source=1&gclid=CjwKCAjwzN-vBhAkEiwAYiO7oC9iLlxJS-8geSn-7JKSXzq410j9qHZYHxv4nD1MSpbUWNjJHiBTjhoC9A4QAvD_BwE
Jiskani IM, Silva JMND, Chalgri SR, Behrani P, Lu X, Manda E (2020) Mine health and safety: influence of psychosocial factors on musculoskeletal disorders among miners in Pakistan. Int J Min Min Eng 11(2):152–167. https://doi.org/10.1504/IJMME.2020.108646
Pavlak A (2004) Project troubleshooting: tiger teams for reactive risk management. Pro Manag J 35(4):5–14. https://doi.org/10.1177/875697280403500403
Akcin H (2021) A GIS-based building risk assessment for the subsidence due to undercity coal mining activities in Zonguldak. Turkey Arab J Geosci 14(5):1–13. https://doi.org/10.1007/s12517-021-06702-6
Karahan V, Akosman C (2018) Occupational health risk analysis and assessment in cement production processes. Turk J Sci Technol 13(2):29–37
Bartlett PJ (2011) Strategic and tactical risk in mining. Proceedings of the Fourth International Seminar on Stratejik versus Tactical Approaches in Mining. Australian Centre for Geomechanics, Perth. 349–357. https://doi.org/10.36487/ACG_rep/1108_28_Bartlett
Joy J (2004) (2004) Occupational safety risk management in Australian mining. Occup Med 54:311–315. https://doi.org/10.1093/occmed/kqh074
Ott T, Swiaczny F (2001) Time-integrative geographic information systems –management and analysis of spatio-temporal data. Springer. https://doi.org/10.1007/978-3-642-56747-6
Christakos G, Bogaert P, Serre M (2002) Temporal GIS: advanced functions for field-based application. Springer-Verlag, Berlin Heidelberg New York. https://doi.org/10.1007/978-3-642-56540-3
Ahola T, Virrantaus K, Krisp JM, Hunter GJ (2007) A spatio-temporal population model to support risk assessment and damage analysis for decision-making. Int J Geogr Inf Sci 21(8):935–953. https://doi.org/10.1080/13658810701349078
Feng J, Dong Y, Song L (2016) A spatio-temporal analysis of urban crime in Beijing: Based on data for property crime. Urban Stud 53(15):3223–3245. https://doi.org/10.1177/0042098015612982
Mondal B, Sharma P, Kundu D, Bansal S (2021) Spatio-temporal assessment of landscape ecological risk and associated drivers: a case study of Delhi. Environ and Urban ASIA 12(1):85–106. https://doi.org/10.1177/09754253211007830
Shao H, Sun X, Lin Y, Xian W, Zhou Y, Yuan L, Qi J (2021) A method for spatio-temporal process assessment of eco-geological environmental security in mining areas using catastrophe theory and projection pursuit model. Prog Phy Geo: Earth and Environ 45(5):647–668. https://doi.org/10.1177/0309133320982542
Li C, Zhang X, Liu X (2012) Mine safety information technology in the framework of digital mine. Safety Sci 50(4):846–850. https://doi.org/10.1016/j.ssci.2011.08.022
Wu Y, Chen M, Wang K, Fu G (2019) A dynamic information platform for underground coal mine safety based on internet of things. Safety Sci 113(3):9–18. https://doi.org/10.1016/j.ssci.2018.11.003
Wu B, Wang J, Zhong M et al (2023) Multidimensional analysis of coal mine safety accidents in china – 70 years review. Min Met Explor 40:253–262. https://doi.org/10.1007/s42461-022-00722-w
Allen MJ, Allen RT, Davis C, McLeod G (2021) Exploring spatial patterns of Virginia tornadoes using kernel density and space-time cube analysis. ISPRS Int J Geo-Inf 10(5):310. https://doi.org/10.3390/ijgi10050310
Jacquez MG, Greiling AD, Kaufmann MA (2005) A. design and implementation of a space-time intelligence system for disease surveillance. J Geograph Syst 7:7–23. https://doi.org/10.1007/s10109-005-0147-6
Mountrakis G, Gunson K (2009) Multi-scale spatiotemporal analyses of moose-vehicle collisions: a case study in northern Vermont. Int J Geogr Inf Sci 23(11):1389–1412. https://doi.org/10.1080/13658810802406132
Nakaya T, Yano K (2010) Visualising crime cluster in a space-time cube: an exploratory data-analysis approach using space-time kernel density estimation and scan statistics. Transaction in GIS 14(3):223–239. https://doi.org/10.1111/j.1467-9671.2010.01194.x
Gonschorek J, Tyrallova L (2012) Geovisualization and geostatistics: a concept for the numerical and visual analysis of geographic mass data. International Conference on Computational Science and Its Applications. Springer, Berlin, Heidelberg, pp 208–219
Morales BR, Varela ERD, Marey-Perez MF (2013) Spatiotemporal analysis of vehicle collisions involving wild boar and roe deer in NW Spain. Accident Anal & Prevention 60:121–133. https://doi.org/10.1016/j.aap.2013.07.032
Walker BB, Schuurman N, Hameed SM (2014) A GIS-based spatiotemporal analysis of violent trauma hotspots in Vancouver, Canada: identification, contextualisation and intervention. BMJ Open 4:e003642. https://doi.org/10.1136/bmjopen-2013-003642
Ye X, Xu X, Lee J, Zhu X, Wu L (2015) Space-time interaction of residential burglaries in Wuhan, China. Applied Geogr 60:210–216. https://doi.org/10.1016/j.apgeog.2014.11.022
Song J, Wen R, Yan W (2018) Identification of traffic accident clusters using Kulldorff’s space-time scan statistics. In 2018 IEEE International Conference on Big Data (Big Data), p.3162–3167
Kang Y, Cho N, Son S (2018) Spatiotemporal characteristics of elderly population’s traffic accidents in Seoul using space-time cube and space-time kernel density estimation. PLoS ONE 13(5):e0196845. https://doi.org/10.1371/journal.pone.0196845
Yıldırım V, Yurdakul E, Karaağaç GA, Koçer M, Uyguçgil H (2023) Spatiotemporal analysis of traffic accidents in Eskişehir City Center. Turk J Remote Sens GIS. 4(1):17–32. https://doi.org/10.48123/rsgis.1167844
Mo C, Tan D, Mai T, Pang W, Zhang Z, Bei C, Qin J (2020) An analysis of spatiotemporal pattern for COVID-19 in China based on space-time cube. J Med Virol Wiley 92(9):1587–1595. https://doi.org/10.1002/jmv.25834
Andrade LA, Gomes DS, Lima SVMA, Duque AM, Melo MS, Goes MAO, Ribeiro CJN, Peixoto MVS, Souza CDF, Santos AD (2020) COVID-19 mortality in an area of northeast Brazil: epidemiological characteristics and prospective spatiotemporal modelling. Epidemiol Infect 148(e288):1–7. https://doi.org/10.1017/S0950268820002915
Jing C, Dong M, Du M, Zhu Y, Fu J (2020) Fine-grained spatiotemporal dynamics of inbound tourists based on geotagged photos: a case study in Beijing, China. IEEE Access 8:28735–28745. https://doi.org/10.1109/ACCESS.2020.2972309
Yi S, Wang H, Yang S, Xie L, Goa Y, Ma C (2021) Spatial and temporal characteristics of hand-foot-and-mouth disease and its response to climate factors in the Ili River Valley Region of China. Int J Enviro Res Publ Health 18(4):1954. https://doi.org/10.3390/ijerph18041954
Makovska I, Howard MW, Bezymennyi M, Witkowski L, Nedosekov V, Kornienko L, Karpulenko M, De Nardi M (2021) Descriptive spatiotemporal analysis of rabies in domestic carnivores and wildlife in Ukraine in 2012–2018. Med Weter 77(11):546–562. https://doi.org/10.21521/mw.6589
Buckley A (2018) Working with temporal data in ArcGIS. Available from: https://www.esri.com/about/newsroom/arcuser/working-with-temporal-data-in-arcgis/. Accessed 15 Nov 2022
Ferreira KR, Oliveira AG, Monteiro AMV, Almeida DBFC (2015) Temporal GIS and spatiotemporal data sources. Proceedings Brazilian symposium on geoinformatics XVI GEOINFO, Campos do Jordao, Brazil. p. 1–13
Pak U, Ma J, Ryu U, Ryom K, Juhyok U, Pak K, Pak C (2020) Deep learning-based PM25 prediction considering the spatiotemporal correlations: a case study of Beijing, China. Sci Total Environ 699:133561. https://doi.org/10.1016/j.scitotenv.2019.07.367
Ren M, Cheng G, Zhu W, Nie W, Guan K, Yang T (2022) A prediction model for surface deformation caused by underground mining based on spatio-temporal associations. Geomat Nat Hazards Risk 13(1):94–122. https://doi.org/10.1080/19475705.2021.2015460
Alkan M, Cömert Ç (2010) A design of temporal geographic information systems (TGIS) for Turkish land register and cadastre data. Sci Res Essays 5(7):700–708 (https://hdl.handle.net/20.500.12628/1948)
International Standard Organization (ISO), 8601 (2004) Data elements and interchange formats - Representation of dates and times. Switzerland, Geneva
Xiao W, Xu J, Lv X (2019) Establishing a georeferenced spatiotemporal database for Chinese coal mining accidents between 2000 and 2015. Geomat Nat Hazards Risk 10(1):242–270. https://doi.org/10.1080/19475705.2018.1521476
Gupta, R.P., 2005. Remote Sensing Geology, second ed. Springer International Edition, 627.
Paramasivam CR, Venkatramanan S (2019) An Introduction to Various Spatial Analysis Techniques. Chapter 3 in book: GIS and Geostatistical Techniques for Groundwater Science. Publisher: Elsevier, pp 23–30. https://doi.org/10.1016/B978-0-12-815413-7.00003-1
Goodchild MF (2018) Spatial data analysis. Chapter in book: Encyclopedia of database systems. Publisher Springer, New York, pp 3570–3575. https://doi.org/10.1007/978-1-4614-8265-9
O’Shea TC, Nicholls K (2003) Police crime analysis: a survey of US police departments with 100 or more sworn personnel. Police Pract Res 4(3):233–250. https://doi.org/10.1080/1561426032000113852
Gürer C (2004) Suç analizi: çağdaş polisin vazgeçilmez silahı (Crime analysis: the indispensable weapon of the modern police). Çağın polisi dergisi (Police of the age magazine: in Turkish). 34 (3). Available via DIALOG. https://www.caginpolisi.com.tr/suc-analizi-cagdas-polisin-vazgecilmez-silahi/
Gottlieb S (1998) Crime analysis; from first report to final arrest, California: Alpha Publising. 596p. Available via DIALOG. https://www.ojp.gov/ncjrs/virtual-library/abstracts/crime-analysis-first-report-final-arrest
Jones MC, Marron JS (1996) A brief survey of bandwidth selection for density, estimation. J Am Stat Assoc 91(433):401–407
ESRI a (2024) ArcGIS Pro resources, hot spot analysis (Getis-Ord Gi*) (spatial statistics). Available via DIALOG. https://pro.arcgis.com/en/pro-app/3.1/tool-reference/spatial-statistics/hot-spot-analysis.htm
ESRI b (2024) ArcGIS Pro Resources, What is a z-score? What is a p-value? Available via DIALOG. https://pro.arcgis.com/en/pro-app/latest/tool-reference/spatial-statistics/what-is-a-z-score-what-is-a-p-value.htm
Chainey S (2021) Understanding crime: analyzing the geography of crime. Esri Press. ISBN 9781589485846, eISBN 9781589485853
Piza EL, Baughman JH (2021) Modern policing using ArcGIS Pro. Esri Press. ISBN 9781589485976, eISBN 9781589485983.
Li Y, Zhang L, Yan J, Wang P, Hu N, Cheng W, Fu B (2017) Mapping the hotspots and coldspots of ecosystem services in conservation priority setting. J Geogr Sci 27(6):681–696. https://doi.org/10.1007/s11442-017-1400-x
ESRI c (2024) ArcGIS Pro resources. Crime analysis tool reference. Available via DIALOG. https://doc.arcgis.com/en/arcgis-solutions/11.2/reference/tool-reference.htm
Kyriakidis PC, Journel AG (1999) Geostatistical space-time models: a review. Math Geol 31(6):651–684. https://doi.org/10.1023/A:1007528426688
Kveladze I, Kraak MJ, Van Elzakker CP (2015) The space-time cube as part of a GeoVisual analytics environment to support the understanding of movement data. Int J Geogr Inf Sci 29(11):2001–2016. https://doi.org/10.1080/13658816.2015.1058386
Yue S, Pilon P, Phinney B, Cavadias G (2002) The influence of autocorrelation on the ability to detect trend in hydrological series. Hydro Proces 16:1807–1829. https://doi.org/10.1002/hyp.1095
ESRI d (2024) ArcGIS Pro Resources, How emerging hot spot analysis works. Available via DIALOG. https://pro.arcgis.com/en/pro-app/2.8/tool-reference/space-time-pattern-mining/learnmoreemerging.htm
European Commission Directorate-General for International Partnerships (EuropeAid HQ) (2021) Capacity building of occupational physicians for the assessment and prevention of psychosocial risks in mining sector. Available via DIALOG. https://www.developmentaid.org/organizations/awards/view/205017/capacity-building-of-occupational-physicians-for-the-assessment-and-prevention-of-psychosocial-risks
Işık İ, Öz Aktepe Ş, Çetin Özbudak E, Ceylan F, Kuzdağ Y, Dönmez A (2022) Psychosocial Security studies in coal mines: qualitative evaluation from the perspective of mining engineers and occupational safety experts (in Turkish). Social Security (20):592–622. https://doi.org/10.21441/sosyalguvence.1026839
Acknowledgements
The authors thanks Esri Türkiye, Turkish Hard Coal Enterprise, Kozlu Enterprise officials, Plan Bureau and Occupational Health and Safety Branch Office staff, Serkan Sarginoğlu, and Alaaddin Cakir for their contributions to greatly improve the paper.
Funding
Open access funding provided by the Scientific and Technological Research Council of Türkiye (TÜBİTAK).
Author information
Authors and Affiliations
Contributions
All authors contributed to the study conception and design. The materials were prepared by all authors. Analyses were performed by Pınar Eksert and Hakan Akcin. The first draft of the article was written by Hakan Akcin and was developed with the contributions of all authors. All authors have read and approved the final manuscript.
Corresponding author
Ethics declarations
Conflict of Interest
The authors declare no competing interests.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Eksert, P., Akcin, H. A Review on the 3D Cartographic and Spatiotemporal GIS Models for Safety of Accidents in Deep Underground Coal Mines. Mining, Metallurgy & Exploration 41, 1221–1243 (2024). https://doi.org/10.1007/s42461-024-00977-5
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s42461-024-00977-5