ALBERT

Description: Paleoclimate reconstruction based on assimilation of proxy observations requires specification of the control variables and their background statistics. As opposed to numerical weather prediction (NWP), which is mostly an initial condition problem, the main source of error growth in deterministic Earth system models (ESMs) regarding the model low-frequency response comes from errors in other inputs: parameters for the small-scale physics, as well as forcing and boundary conditions. Also, comprehensive ESMs are non-linear and only a few ensemble members can be run in current high-performance computers. Under these conditions we evaluate two assimilation schemes, which (a) count on iterations to deal with non-linearity and (b) are based on low-dimensional control vectors to reduce the computational need. The practical implementation would assume that the ESM has been previously globally tuned with current observations and that for a given situation there is previous knowledge of the most sensitive inputs (given corresponding uncertainties), which should be selected as control variables. The low dimension of the control vector allows for using full-rank covariances and resorting to finite-difference sensitivities (FDSs). The schemes are then an FDS implementation of the iterative Kalman smoother (FDS-IKS, a Gauss–Newton scheme) and a so-called FDS-multistep Kalman smoother (FDS-MKS, based on repeated assimilation of the observations). We describe the schemes and evaluate the analysis step for a data assimilation window in two numerical experiments: (a) a simple 1-D energy balance model (Ebm1D; which has an adjoint code) with present-day surface air temperature from the NCEP/NCAR reanalysis data as a target and (b) a multi-decadal synthetic case with the Community Earth System Model (CESM v1.2, with no adjoint). In the Ebm1D experiment, the FDS-IKS converges to the same parameters and cost function values as a 4D-Var scheme. For similar iterations to the FDS-IKS, the FDS-MKS results in slightly higher cost function values, which are still substantially lower than those of an ensemble transform Kalman filter (ETKF). In the CESM experiment, we include an ETKF with Gaussian anamorphosis (ETKF-GA) implementation as a potential non-linear assimilation alternative. For three iterations, both FDS schemes obtain cost functions values that are close between them and (with about half the computational cost) lower than those of the ETKF and ETKF-GA (with similar cost function values). Overall, the FDS-IKS seems more adequate for the problem, with the FDS-MKS potentially more useful to damp increments in early iterations of the FDS-IKS.

Description: In climate reanalyses for multi-decadal or longer scales with coupled atmosphere-ocean General Circulation models (CGCMs) it can be assumed that the growth of prediction errors arises chiefly from imprecisely known model parameters, which have a nonlinear relationship with the climate observations (paleoclimate proxies). Also, high-resolution CGCMs for climate analysis are extremely expensive to run, which constrains the applicability of assimilation schemes. In a model framework where we assume that model dynamic parameters account for (nearly) all forecast errors at observation times, we compare two computationally efficient iterative schemes for approximate nonlinear model parameter estimation and joint flux estimation (taking the specific shape of freshwater from melting in the Greenland ice sheet), and its physically consistent state. First, a trivial adaptation of the strong constraint incremental 4D-Var formulation leads to what we refer to as the parameter space iterative extended Kalman smoother (pIKS); a Gauss-Newton scheme. Second, a so-called parameter space fractional Kalman smoother (pFKS) is an alternative controlled-step line search, which can potentially be a more stable approach. While these iterative schemes have been used in data assimilation, we revisit them together within the context of parameter estimation in climate reanalysis, as compared to the more general 4D-Var formulation. Then, the two schemes are evaluated in numerical experiments with a simple 1D energy balance model (Ebm1D) and with a fully-coupled Community Earth System Model (CESM v1.2). Firstly, with Ebm1D the pFKS obtains a cost function similar to the adjoint method with highly reduced computational cost, while an ensemble transform Kalman filter with an m = 60 ensemble size (ETKF60) behaves slightly worse. The pIKS behaves worse than the ETKF60, but an ETKF10 (m = 10) is even worst. Accordingly, with CESM we evaluate the pKFS and the ETKF60 along with an ETKF with Gaussian Anamorphosis (ETKF-GA60). From all the options, the pFKS has the lowest cost function and seems the favored overall option under heavy computational restrictions, but the ETKF obtains better estimates of the flux term.