JWG C.8: Methodology of comparing/validating climate simulations with geodetic data

Chair: Jürgen Kusche (Germany)
(Affiliation: Commisison 2, ICCT)

Introduction

Climate model simulations provide a unique tool for understanding past and present climates, and projecting future climate under given scenarios. Comparing simulations to observed data is of key importance for understanding uncertainties and systematic errors, attributing causal relations among competing hypotheses, and lending confidence to predictions. One way geodesy contributes to climate sciences is via providing a unique data record for comparing, validating and testing climate model simulations, e.g. with respect to observed sea level change, mass redistribution or water vapor timeseries. However, what is required from geodesy is beyond ‘just’ providing long and stable data sets, and challenges must be addressed that require revisiting the methodology.

Climate model simulations often consist of large, single- or multi-model ensembles that need to be evaluated as a whole (i.e. not just the ensemble mean which averages out real variability). Model simulations encompass hundreds of  dynamically linked variables with increasing spatial and temporal resolution, as e.g. in climate monitoring applications. Simulations are prone to drifts and biases, and climate science is not only interested in ‘deterministic’ (or mean) outcomes but e.g. in how statistics of extreme events or turbulent phenomena like eddy kinetic energy varies over time. Testing climate models e.g. for the anthropogenic fingerprint requires that such statistics are derived from (geodetic) data, where uncertainties must be assigned that characterize both measurement and sampling errors. This will require that geodetic methods are critically reviewed and new approaches e.g. that deal with large data quantities and ensembles and/or the reconstruction of data gaps, and that capture statistics beyond the mean and RMS be developed. Moreover, as geodetic techniques evolve (and are integrated with non-geodetic observables), new climate observables as e.g. Earth’s Energy Budget come into reach and new methods must be developed.

Objectives

• To contribute to the understanding of past, present and future climate via comparing/validating/testing climate models with geodetic data
• To promote and advance the development of rigorous methods across climate science and geodesy
• To safeguard the proper evaluation of the complex geodetic data sets, which are often built from multiple techniques, over overlapping or non-overlapping time periods, in the presence of background model and other analysis technique changes, in the presence of instrumental errors and biases and the intricacies of geodetic reference frames (i.e. develop ‘best-practice’ examples)
• To facilitate communication of climate scientists and geodesists with respect to methodological issues
• To enable the reanalysis of long geodetic data sets

Program of Activities

• The WG will first seek to stimulate exchange and collaboration across geodesists from different background (atmospheric remote sensing, altimetry, GRACE) and climate scientists.
• It will organize splinter meetings along conferences and promote session proposals e.g. at EGU and AGU conferences.
• It will eventually seek to consolidate the geodesy methodology and disseminate results in form of a white paper / review paper.

Members

• Henryk Dobslaw (Germany)
• Petra Friedrichs (Germany)
• Vincent Humphrey (USA)
• Laura Jensen (Germany)
• Anna Klos (Poland)
• Felix Landerer (USA)
• Ben Marzeion (Germany)
• Anne Springer (Germany)