The primary goal of this thesis is to accurately map the geology of the East Rand Basin by creating a 3D model. This thesis shows that these methods can be used for fundamental geological research, if the data are of the same quality and quantity.
Introduction
- Introduction
- Problem statement
- Research aims and objectives
- Study layout
This will help in visualizing the geology and geological features of the Eastern Basin. Therefore, ArcGIS will be used in this area to help produce an accurate 3D geological model of the Eastern Basin.
Literature review
Introduction
Background of the East Rand and the East Rand Basin
Blesbokspruit One of the two main streams that flow through the East Rand Basin area. Rietspruit One of the two main streams that flow through the East Rand Basin area.
Mining activity within the East Rand Basin
- Shaft mining / Shaft sinking
- Opencast mining / Open-pit mining
- The impact of the mining methods on the East Rand Basin
This mining took place on top of the Witwatersrand rocks (Brakpan) and on top of the dolomites (Springs) within the East Rand Basin (GreatMining, 2014). According to DWA (2013a), the water level within the mine cavities of the East Rand basin is gradually rising and will continue to rise until the water is pumped out of the voids.
Geology of the East Rand Basin
In the western part of the basin, the Witwatersrand sediments are overlain by the Ventersdorp Supergroup (Scott, 1995). Over the southwestern parts of the basin, Black Reef quartzite is deposited on the Ventersdorp lavas.
The East Rand Basin’s lithostratigraphic column
Therefore, the depth and thickness measurements must be converted from feet to meters before it is used to calculate the average thickness. Where the thickness of SACS and Scott corresponds to the thickness of the western and eastern columns of Antrobus and Whiteside, the average of the four depths is calculated. However, when the readings differ significantly, a minimum and maximum depth is given (such as Black Reef).
The reason for this is that the geological layer is affected by the thinning mentioned under section 2.4. The new table will also illustrate which subgroup, group and supergroup each geological layer belongs to. The Ventersdorp Supergroup is also present in this new table, although it only appears in the western region of the East Rand Basin.
The reason for adding this Supergroup is that the new table is simply an illustration of the average thickness and arrangement of the various geological strata within the East Rand Basin. The methods and materials that will be used to capture the geology of the East Rand Basin within a 3D geological model will be explained and described in the next chapter.
Method and Materials
Introduction
GIS
The reason for this is that the resolution of the data is represented by cells, which makes it difficult to illustrate lines and areas. This can be seen by comparing the line feature of Figure 3.1 and Figure 3.2: the larger the cells, the lower the resolution. Most of the data will need to be converted as it is in vector data format.
According to ESRI (2013), a triangular irregular network surface (TIN) consists of vector-based digital geographic data constructed by triangulating a series of points (vertices). This type of Kriging is known to be linear because the estimates are weighted linear combinations of the available data. When it comes to the semivariogram (Figure 3.6), the other Kriging methods calculate the semivariogram based on known data locations and use this semivariogram to make predictions for the unknown locations, while the Empirical Bayesian Kriging method takes into account the error introduced by estimating the underlying semivariogram.
When it comes to raster production, EBK is known to be slower than other methods. Online 3D Terrain Visualization - Most of the applications used for 3D terrain visualization are known to be open source web-based systems. A 3D model, according to Wainwright & Mulligan (2004), is known as an abstraction of reality that represents a complex reality in the simplest way that is adequate for the purpose of modeling.
Creating the models
- ArcGIS software
- Data collection
- Geodatabase development for the East Rand Basin
- Building the 3D models
The East Rand catchment generic feature class will consist of the generic features used to generate the models. These features include the East Rand Basin outline and the New Area polygon. Model A is divided into A(a), A(b) and A(c), so that the completed models A(b) and A(c) will be used as submodels in Model B and Model C.
Models A(a), B and C will be created from the same borehole data (National Groundwater Archive Geodatabase), while Model A(b) will be created from average depths and thicknesses (SACS and the new lithostratigraphic column) and Model A(c) ) will be created from the DEM surfaces provided by the North-West University. Using the 3D Analyzer tools, the top and bottom depth, in meters, of each geological layer at the various borehole locations (see Figure 3.14), will be converted to 3D surface features. After these TINs are created, the "Edit TIN Tool" will be used to clip the top and bottom TINs to the East Rand basin's polygon feature class.
The East Rand Basin polygon feature class will be used for each of these geological layers, due to the fact that they will only be seen as "basic". After creating these DEMs, each DEM will be converted to a TIN using the "Raster To TIN Tool". After these TINs are created, the TINs will go through the same processes/steps as under Model A (a), but instead of the East Rand Basin polygon feature, the New Area polygon feature will be used to extrude between the TINs.
Results
Introduction
Data restrictions
Challenges with the data were encountered when it came to working with the different depths, because the depths provided by the National Groundwater Archive Geodatabase are an indication of the distance from the surface and not the distance from sea level. These columns usually indicate where each group and subgroup is located from sea level, while the borehole data gives the distance from the surface. This then leads to the study of topographic maps in order to determine the surface height of the study area above sea level.
Knowing this elevation will make it easier to match the geologic layers of the borehole data with the stratigraphic columns. The deepest borehole reaches a depth of -351 m and the depth to which the model must be built is that of the Main Reef formation, which reaches a depth of -2,330 m. The DEMs of the Main Reef, Black Reef and Kimberley Reef provided by North West University were found to be accurate, although they only provided a single depth – the bottom depth – and not both the top and bottom depths .
The depths derived by the SACS Task Group were found to be accurate, but it also presented some limitations to the study due to the fact that an average depth is given for each geological layer. To eliminate this, point data must be created for each of these geological layers. This type of data should be in the same format, containing x, y and z data, as those in the National Groundwater Archive Geodatabase.
Results of the ArcGIS spatial analysis methods
- Model A
- Model B
- Model C
- Comparing the models with each other and with geological cross-sections
The problem with this is, as mentioned under 4.2., that the entire layer is one solid block with no changes in depth. This can be seen in Table 4.3 above, where the upper surfaces of the main ridge appear darker (view N-N) than the smoother lower surfaces of the main ridge (view N-S). After shearing, the top depth TINs were found to be smaller than the bottom depth TINs (see Table 4.4), but this did not affect the results of each layer.
Comparing the north-south profile plots (Figures 4.6 and 4.7) of the lower and upper depth TINs produced by the different models revealed no similarities between the two. These available cross-sections are compared below with the profile charts of the different reefs. By comparing the cross section ABC of DWA (Figure 4.14) with the profile plots of the different reefs (Figure 4.15), these layers can be said to be quite accurate, as all layers show the same concave uplift character.
The Black and Main Reefs' profile graphs also indicate the same climax of the Springs Monocline, which is illustrated by the Black and Main (Nigel) Reef in Figure 4.14. Below, the cross-section AB (Figure 4.16) designed by Pitts (1990) is compared with the profile graphs of the various reefs within the East Rand Basin (Figure 4.17). The Kimberley Reef's profile graph shows no agreement with the two diameters and therefore cannot be said to be accurate in terms of the entire East Rand basin.
Conclusion and recommendations
Conclusion
Along with the above information, it should be kept in mind that the three reefs were found to be accurate, as they showed the same concave uplift character when compared to the DWA cross-sections ABC (Figure 4.14). In addition, the method for Model C is judged to be the most accurate of the three methods based on self opinion; Unlike a TIN that just connects the dots, Model C was created using the Kriging method. This method predicted the values of points for which no information existed, based on the data from the sample points with information.
Through this process, the Kriging method also tried to get the error or mean residual close to or equal to 0.
Recommendations
Spatial data modeling for 3D GIS. http://books.google.co.za/books?id=X6a2f75ky3gC&printsec=frontcover&source=gbs_ge_sum mary_r&cad=0#v=onepage&q&f=false Access date: 15 November 2013. X&ei=eNgaVPv9SmoEv0Dw7 page&q=korte&f=false Access date: 15 November 2013. Central and East Rand Gold Mine Flooding: An Investigation into Flow Rate Controls, Water Quality and Anticipated Impacts of Flooded Mines.
The Drainage of the Far Eastern Rand Mining Basin: A Critical Appraisal of the Government's Approach to Addressing Associated Environmental Problems. Threats and opportunities for post-closure development in 71 dolomite gold mining areas in the West Rand and Far West Rand (South Africa) – a hydraulic perspective Part 1: Mining legacy and future threats.
