heteroskedastic

Esper stated that the variance in Gotland was "heteroskedastic" and they did exercised to eliminate the supposed heteroskedasticity by a power-transformation by a power of 0.32.

Log-Normality in Gotland « Climate Audit 2005

Ah, actually Google tells me the spelling is 'heteroskedastic'.

Lovely sunny sunshine accommodatingly 2004

Footnote for those who care: for temperature trend estimation, bounding or testing, I think that you need to use something like Engel's ARCH LM test "autoregressive conditional heteroskedastic" Lagrange Multiplier, which is based on the regression of the squared residuals on the lagged squared residuals, for as many lags you want to include, and which the test statistic equal to the sample size times the R-squared is approximately a Chi-square.

USCCSP: Temperature Trends in the Lower Atmosphere « Climate Audit 2006

And third if the answer is the series, could you explain why homoskedastic is better than heteroskedastic for an individual series?

Wilson on Yamal Substitution « Climate Audit 2006

I thought that our comment about heteroskedastic-autocorrelation consistent standardization was pretty neat.

Reply to Huybers #1 « Climate Audit 2005

These issues are well-known in econometrics, where they have led to “heteroskedastic-autocorrelation consistent” estimators.

Spurious #5: Variance of Autocorrelated Processes « Climate Audit 2005

In our Reply, we calculated variances using a “heteroskedastic-autocorrelation consistent” variance estimator from econometrics.

Reply to Huybers #3: Principal Components « Climate Audit 2005

I also want to mention in passing mathworld.wolfram.com , which is where I run to when someone says something like ‘heteroskedastic’.

Log-Normality in Gotland « Climate Audit 2005

Newey-West (as referenced in Stata) isn't GLS at all; it's a correction to the standard errors for autocorrelated / heteroskedastic noise when using

RealClimate 2009

If you use the White correction to least squares, it's a very restrictive form of GLS which assumes the noise is uncorrelated (although heteroskedastic), so the variance-covariance matrix is not the identity matrix but is still diagonal.

RealClimate 2009