TA的每日心情 | 奋斗 2016-6-30 16:07 |
---|
签到天数: 90 天 [LV.6]常住居民II
|
发表于 2014-2-20 17:19:26
|
显示全部楼层
从图书馆借来了《Geostatistical Reservoir Modeling》这本书,从Chapter 4开始看起(书实在是太枯燥了,请原谅我肿么跳着断断续续看),看到了Coordinate Transformation这个小标题下面的一些文字,觉得和作者君上面说的关于“数学变换”的一些内容正好能够贴合上,所以将下面这段文字摘录进来:
The coordinate transformations presented in Chapter 3 are necessary before variogram calculation. In presence of vertical wells, the vertical variogram does not depend on the stratigraphic coordinate transform, as long as calculations are limited to data within the appropriate stratigraphic layer and facies type. The horizontal variogram, however, is very sensitive to the stratigraphic Zrel coordinate transform. Attempting variogram calculation prior to such stratigraphic coordinate transformation can lead the modeler to the erroneous concousion that the data have no horizontal correlation. Data sparsity may also lead to the same erroneous conclusion.
——摘自《Geostatistical Reservoir Modeling》P108
可想而这,在变差函数调整之前,将数据进行一定程度的数学变换是非常必要滴——在我之前做建模的过程中,还有周围的好多人,跳过Transformation直接开始进行变差函数的调整,肿么做实际上是不科学的。
期待Joseph关于“闲扯地质统计学”的后续文章~
|
|