Estimation of global temperature trends: What's important and what isn't

In this essay we discuss the development of and estimation of uncertainties in the global surface temperature record. We briefly discuss the similarities in and differences between the records from the institutions that produce such series. We then consider the numerous issues that must be addressed to enable accurate estimates to be derived. We consider these in their order of importance with respect to the record: biases in the sea surface temperature data, exposure of land-based thermometers before about 1900, urbanization effects in some series, and, finally, the homogeneity of individual land-based records.