The boundary line, as described by (Shatar and McBratney 2004), is a 4th-degree polynomial. The problem arises when the input data has been extrapolated via regression and/or kriging; where data at the extremities may become inaccurate. This can cause the boundary line to become extremely low such that the input variable became, wrongly, the yield-limiting factor.
Solution:
tl;dr? Change how the boundary line is fit.
During discussions, various techniques have been discussed to overcome the problem. The following is a list of the discussed techiques.
Problem:
The boundary line, as described by (Shatar and McBratney 2004), is a 4th-degree polynomial. The problem arises when the input data has been extrapolated via regression and/or kriging; where data at the extremities may become inaccurate. This can cause the boundary line to become extremely low such that the input variable became, wrongly, the yield-limiting factor.
Solution:
tl;dr? Change how the boundary line is fit.
During discussions, various techniques have been discussed to overcome the problem. The following is a list of the discussed techiques.