[Paper Review] Planetary boundary layer depth in Global climate models induced biases in surface climatology

Richard Davy, Igor Ezau|arXiv (Cornell University)|Sep 30, 2014

Climate variability and models20 references3 citations

TL;DR

This study demonstrates that biases in planetary boundary layer (PBL) depth representation within global climate models (GCMs) are a primary cause of discrepancies between simulated and observed surface temperature trends and variability—accounting for up to 60% of the difference in trends and 50% in variability in CMIP5 models. The authors show that inaccurate PBL depth parameterization, especially in stably stratified Arctic conditions, fundamentally limits model fidelity in simulating surface climate responses to forcing changes.

ABSTRACT

The Earth has warmed in the last century with the most rapid warming occurring near the surface in the arctic. This enhanced surface warming in the Arctic is partly because the extra heat is trapped in a thin layer of air near the surface due to the persistent stable-stratification found in this region. The warming of the surface air due to the extra heat depends upon the amount of turbulent mixing in the atmosphere, which is described by the depth of the atmospheric boundary layer (ABL). In this way the depth of the ABL determines the effective response of the surface air temperature to perturbations in the climate forcing. The ABL depth can vary from tens of meters to a few kilometers which presents a challenge for global climate models which cannot resolve the shallower layers. Here we show that the uncertainties in the depth of the ABL can explain up to 60 percent of the difference between the simulated and observed surface air temperature trends and 50 percent of the difference in temperature variability for the Climate Model Intercomparison Project Phase 5 (CMIP5) ensemble mean. Previously the difference between observed and modeled temperature was thought to be largely due to differences in individual models treatment of large-scale circulation and other factors related to the forcing, such as sea-ice extent. While this can be an important source of uncertainty in climate projections, our results show that it is the representation of the ABL in these models which is the main reason global climate models cannot reproduce the observed spatial and temporal pattern of climate change. This highlights the need for a better description of the stably-stratified ABL in global climate models in order to constrain the current uncertainty in climate variability and projections of climate change in the surface layer.

Motivation & Objective

• To investigate the role of planetary boundary layer (PBL) depth representation in global climate models (GCMs) as a source of bias in surface temperature simulations.
• To quantify how uncertainties in PBL depth contribute to the discrepancy between observed and modeled surface air temperature trends and variability in CMIP5 models.
• To challenge the prevailing assumption that differences in large-scale circulation or forcing factors (e.g., sea-ice extent) are the dominant causes of model-observation discrepancies.
• To highlight the need for improved representation of stably stratified boundary layers in GCMs to reduce uncertainty in climate projections.
• To provide evidence that PBL depth is a critical, previously underappreciated factor in surface climatology biases across global climate models.

Proposed method

• Analysis of the CMIP5 ensemble mean and individual model outputs to compare simulated surface air temperature trends and variability with observational data.
• Use of observational datasets to define reference surface temperature trends and variability, particularly in the Arctic region.
• Sensitivity analysis of PBL depth parameterizations in GCMs to assess their impact on simulated surface temperature responses.
• Comparison of model-simulated PBL depths with observed or reanalysis-based estimates to identify systematic biases.
• Statistical attribution of model-observation differences in temperature trends and variability to PBL depth uncertainty using regression and variance decomposition.
• Focus on stably stratified conditions in the Arctic, where PBL depth is shallow and highly sensitive to turbulent mixing and thermal stability.

Experimental results

Research questions

• RQ1 To what extent does inaccurate planetary boundary layer depth in GCMs contribute to biases in simulated surface air temperature trends?
• RQ2 How much of the observed difference between CMIP5 model outputs and observational data can be attributed to PBL depth misrepresentation?
• RQ3 Why do global climate models fail to reproduce the observed spatial and temporal patterns of climate change, particularly in the Arctic?
• RQ4 How does the depth of the stably stratified planetary boundary layer influence the surface temperature response to climate forcing?
• RQ5 What role does PBL depth uncertainty play relative to other factors such as large-scale circulation or sea-ice extent in driving model-observation discrepancies?

Key findings

• Uncertainties in planetary boundary layer (PBL) depth explain up to 60% of the difference between simulated and observed surface air temperature trends in CMIP5 models.
• PBL depth biases account for approximately 50% of the discrepancy in surface temperature variability between models and observations.
• The study identifies PBL depth as the dominant factor behind surface climatology biases in GCMs, surpassing the influence of large-scale circulation or sea-ice extent in many cases.
• In the Arctic, where surface warming is amplified, the shallow and stably stratified PBL is particularly sensitive to model representation, leading to significant errors in heat trapping and temperature response.
• Models with inaccurate PBL depth parameterization fail to properly simulate the vertical distribution of heat, leading to incorrect surface temperature trends.
• The results suggest that improving the representation of stably stratified boundary layers in GCMs is essential to reducing uncertainty in climate projections and improving model fidelity.