Kalyanjit - 

Land use suitability is not in my repertoire, so I cannot guess very well with what method you are working, but in general, interpolation is likely more accurate than extrapolation. Whatever two points between which you are interpolating, however, will certainly be likely more accurate than the point you are interpolating. I would assume that the point you are to interpolate is multidimensional, and there may be several nonlinear relationships involved. Thus, substantial error may be involved.  The results you have for the situations between which you are interpolating need to be substantially more accurate than required for the results you interpolate.  

Perhaps you could explain your situation more fully for the experts in your field. 

With regard to statistics, do not underestimate the usefulness of graphing results. For comparing continuous data, if that is appropriate, scatterplots may be very useful. And when you can estimate standard errors, do so.  

You could consider a sensitivity analysis where you examine your data to see how far 'wrong' your interpolation could be before causing a substantial practical problem for your study. 

Cheers - Jim