Anselin Luc. Thirty yearsof spatial econometrics[J]. Papers inRegional Science, 2010, 89(1): 3-23.

最后他提到了未来发展的方向与挑战，请解释第二个挑战的意思。尤其是划绿色的部分。

谢谢

A second challenge is to deal with theconceptual questions raised by using ever larger data sets, many of whichresult from automatic data recording. Systems ofsensors, both in the natural and in the social sciences, provide everlarger streams of extremely fine grained data (on a time scale, geographicalscale and individual scale). The standard sample-population paradigm or even thespatial stochastic process paradigm are insufficient to meaningfully addressthe questions raised in the analysis of such massive data sets. Thenotions of equilibrium (like spatial stationarity) on which we rely to developand assess our models are likely to become unrealistic in this context. Also,traditional concepts such as ‘significance’ have little use in the analysis of massivedata sets, since everything tends to be significant. Other ways ofinterpreting and assessing models for such settings need to be developed.Important in this respect is the need for a better understanding of how errors(of measurement, of location, of model specification) propagate through ourinferential systems and bring into question our traditional methods for quantifyinguncertainty.