Time-scale dependent estimation of spatial degrees of freedom

torben.kunz [ at ] awi.de


Effective spatial degrees of freedom (ESDOF) provide a simple metric to characterize the spatial structure of a time-varying field. Methods to estimate the ESDOF from global fields of climate variables are useful in a variety of cases. For example, variance estimation of global averages or field reconstruction, from a limited number ofobservations, always require some knowledge about such basic statistical properties of the underlying processes. However, in many applications the ESDOF estimate depends on the time-scale under consideration. This becomes particularly important in the field of paleo-climate reconstruction, (i) because of the wide range of scales involved in the analysis, and (ii) because the sparsity of climate proxy data increases with time-scale. Here we show how some simple ESDOF measures can be reformulated as a function of frequency. Their properties and interrelations are then illustrated by their application to (a) the Holocene episode of the TraCE-21ka paleo-climate model simulation, and (b) to a simple stochastic-diffusive energy balance model. These model systems indicate a distinct ESDOF reduction with increasing time-scale and, thus, serve as suitable test cases for frequency-dependent ESDOF measures. They also illustrate how additional conclusions, beyond the ESDOF estimate itself, can be drawn from comparison of different measures built on different metrics of the spatio-temporal covariance structure.

Item Type
Conference (Talk)
Primary Division
Primary Topic
Publication Status
Event Details
EGU General Assembly 2018, 07 Apr 2018 - 12 Apr 2018, Vienna, Austria.
Eprint ID
Cite as
Kunz, T. and Laepple, T. (2018): Time-scale dependent estimation of spatial degrees of freedom , EGU General Assembly 2018, Vienna, Austria, 7 April 2018 - 12 April 2018 .

Add to AnyAdd to TwitterAdd to FacebookAdd to LinkedinAdd to PinterestAdd to Email

Geographical region

Research Platforms


Funded by

Edit Item Edit Item