HR: 1340h
AN: OS53A-0981    [Abstracts]
TI: Developing GIS Model for Suitable Sites Selection of Manganese Nodule Development in Clarion-Clipperton Fracture Zone, northeastern Pacific
AU: Ko, Y
EM: ytko@kordi.re.kr
AU: * Park, C
EM: ckpark@kordi.re.kr
AF: Korea Ocean Research and Development Institute, Ansan P.O.Box 29, Seoul 425-600, Korea, Ansan, 426-744, Korea, Republic of
AU: Kim, J
EM: jukim@kordi.re.kr
AF: Korea Ocean Research and Development Institute, Ansan P.O.Box 29, Seoul 425-600, Korea, Ansan, 426-744, Korea, Republic of
AU: Kim, H
EM: hyskim@kordi.re.kr
AF: Korea Ocean Research and Development Institute, Ansan P.O.Box 29, Seoul 425-600, Korea, Ansan, 426-744, Korea, Republic of
AU: Kim, K
EM: kkim@kordi.re.kr
AF: Korea Ocean Research and Development Institute, Ansan P.O.Box 29, Seoul 425-600, Korea, Ansan, 426-744, Korea, Republic of
AB: Manganese nodules are found on the seabed in many areas and have been comparatively well studied for their distribution because of their potential economic importance. The study area is located in the southern central part of the C-C zone, about 2,500 km from the East Pacific Rise (EPR). It consists of three blocks, A2, B2, and C1 between 9¡ãN - 12¡ãN and 128¡ãW - 136¡ãW. The aims of this study are to suggest priority manganese nodule development sites by quantifying the rating values for development potential using probability and statistical methods, and to verify the selection results using three statistical approaches. The collected data were converted into a 100 x 100 m grid using the raster module of the ArcGIS package, and then converted into Data Base File (DBF) data. The each factor considered in this study was then further divided into a few meaningful classes, in order to investigate the dependence of nodule abundance on each class of each factor. As a next step, rating value, average nodule abundance explained by each class, was calculated using the following equation R = ¡Æ w ∙ Np ¨M Nt where R is rating of each class and w is weight of sampling abundance and Np is number of sampling point and Nt is total number of sampling point. Finally, development potential index (DPI) was calculated for each 100 x 100 grid by summing the rating values of all the classes considered. A high DPI indicates a high potential for development and vice versa. The calculated DPIs were verified by comparing them with actual sampling data to understand the most suitable models and the factors that predict the development potential of the study area. Verification was conducted in three ways : success and prediction rates, linear regression analysis, and testing independence. The success rate provides the best model in terms of both area and cumulative frequency diagram for all blocks. The success rate of block A2 provides the best model among the success rates, while the A2toC1 prediction rate provides the best model among prediction rates. The A2 success rate gives the best model out of nine models. The verification for linear regression analysis was conducted to identify the most suitable models. The success rate provides the best model for block A2. The difference in fit between the success and prediction rates is small and both ratings show a general increase irrespective of any anomalies in the data. The B2 success rate model is most suitable for block B2, particularly as its slope is greater than that of the prediction rate and the difference in the R2 value is small. The success rate model was most suitable for block C1. The test of independence was conducted to identify the factors that predict the development potential of an area, by using 21 combinations of the seven factors (copper and nickel grades, slope, aspect, water depth, topography, and transparent layer thickness) for each of blocks A2 and C1 and using 15 combinations of six factors (all but the transparent layer thickness) for block B2. There are no correlations between the factors for blocks A2 and C1, while there are correlations between copper and nickel grade and between topography and water depth in block B2. Verification of the success and prediction rates was conducted for four factors of copper, slope, aspect, and topography in block B2. The results are the same as those using six factors, but the area becomes smaller than when all factors are used.
DE: 0500 COMPUTATIONAL GEOPHYSICS (3200, 3252, 7833)
DE: 0545 Modeling (4255)
DE: 4299 General or miscellaneous
SC: Ocean Sciences [OS]
MN: 2007 Fall Meeting