LIVVkit 2.1: Automated and extensible ice sheet model validation

Abstract

Abstract. A collection of scientific analyses, metrics, and visualizations for robust validation of ice sheet models is presented using the LIVVkit package, version 2.1. This software collection targets stand-alone ice sheet or coupled Earth system models, and handles datasets and operations that require high-performance computing and storage. LIVVkit aims to enable efficient and fully reproducible workflows for post-processing, analysis, and visualization of observational and model-derived datasets in a shareable format, whereby all data, methodologies, and output are distributed to users for evaluation. We demonstrate 5LIVVkit validation for a Greenland ice sheet simulation using the coupled Community Earth System Model, CESM, as well as an idealized stand-alone high-resolution ice sheet model, CISM-Albany. As one example of the capability, LIVVkit analyzes the degree to which models capture the surface mass balance (SMB) and identifies potential sources of bias, using recently available in-situ and remotely sensed data as comparison. Related fields within atmosphere and land surface models, e.g. surface temperature, radiation, and cloud cover, are also diagnosed. Applied to the CESM1.0, LIVVkit identifies a positive 10 SMB bias that is focused largely around Greenland's southwest region that is due to insufficient ablation.

Impact of Grids and Dynamical Cores in CESM2.2 on the Surface Mass Balance of the Greenland Ice Sheet

“…RACMO, version 2.3 11 km (RACMO23; Noël et al., 2015) and version 2.3p2 5.5 km (RACMO2.3p2; Noël et al., 2018, 2019) are RCM simulations targeting Greenland, forced by ERA renalyses products at the domain's lateral boundaries. The RACMO simulations have been shown to perform skillfully against observations and are often used as modeling targets (e.g., Evans et al., 2019; van Kampenhout et al., 2020). The RACMO data sets are used along with the CERES‐EBAF product to validate the radiative fluxes around Greenland.…”

Simulating the Greenland Ice Sheet Surface Mass Balance using an Earth System Model

“…Field data analysis has been carried out through the Land Ice Verification & Validation Toolkit (LIVVkit), an open source software package designed for evaluating ice sheet models and their forcing [Kennedy et al, 2017;Evans et al, 2018]. Three observational SMB datasets available for the GrIS are used: (i) airborne radar, (ii) field accumulation (SMB > 0) measurements, and (iii) field ablation (SMB < 0) measurements [Evans et al, 2018]. The airborne radar data stems from NASA's Operation IceBridge and covers most of the GrIS interior.…”

“…The raw data, as described by Lewis et al [2016], provides seasonal accumulation estimates for a given pixel, uniquely determined by its latitude and longitude. Following Evans et al [2018], a simple time average is applied over all available periods for each record to yield a single accumulation value (in mm w.e. yr 1 ) per location.…”

“…The in-situ field accumulation dataset is a compilation of different field campaigns carried out in the GrIS accumulation zone [Cogley, 2004;Bales et al, 2009;Evans et al, 2018]. Only records that have been retrieved using firn cores, snow pits or stake measurements are included in the evaluation.…”

“…It is important to mention at this point that the spatial coverage of the ablation zone measurements is quite sparse. Indeed, Figure 1 in Evans et al [2018] illustrates that all in-situ ablation data stem from merely 8 transects in total.…”