According to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change, global mean sea level has risen faster since 1900 than in any previous century in the past 3,000 years. This leaves hundreds of coastal cities and millions of people vulnerable to a threat from higher water levels. State-of-the-art climate models provide a crucial means of studying how much and when sea levels will rise. However, to what extent these models are able to represent sea level variations remains an open question. Thus, they need to be evaluated before they can be adopted to predict future sea level changes.
In a recently published article in Atmospheric and Oceanic Science Letters, Dr. Zhuoqi He of the South China Sea Institute of Oceanology led a team to assess the performance of climate models to simulate sea level at low to mid-latitudes around the globe. The results indicated that the models simulated long-term mean sea level relatively well. However, strong biases emerged when the models attempted to reproduce sea level variance. estimated interannual signals over subtropical regions where strong western boundary currents predominate.
“This bias is at least partly due to the misrepresentation of ocean processes due to the relatively low resolution of their historical simulations. We can see that the coastal bias is reduced as the resolution of the model increases,” says He.
“Understanding the causes of model misrepresentation is important for improving model simulation skills, and our study helps in this regard by identifying a direction for future model development to reduce model bias.”
– This press release was provided by the Institute of Atmospheric Physics, Chinese Academy of Sciences website