Week 1:
Good morning!
For my first week, I’ve collected two readings. The first one is called Initial results from Phase 2 of the international urban energy balance model comparison by C. S. B. Grimmond, focusing on the development of urban land surface schemes.
To begin, we need to define a few terms. Land surface models (LSMs) seek to represent the physical and biogeochemical exchanges of energy, water, and other matter that comprise land-atmosphere interactions, especially under conditions of global change. These models originated with simple land-earth biophysics but in the past decade have grown to encompass more interrelated processes. These models measure various flux factors in an environment. Simply defined, flux is “the rate of flow” or the transfer of quantity per unit area. LSMs measure various types of fluxes, including heat flux (transfer of heat between different surfaces and atmospheres), radiative flux (total amount of radiation that is absorbed, converted into heat, and re-emitted), urban flux (which can measure flow of carbon dioxide) and others.
LSMs parametrize energy exchanges between surface and the atmosphere. To model an urban environment, LSMs can vary in their complexity. You can visualize first just a simple slab, then maybe adding different patches of colors to represent different materials, then slowly, like a clay pot taking shape, shapes emerging from the slabs, from simple blobs to complex geometric shapes. The simple slab represents the most simple of LSMs, but then these models can start to take into account 3D geometry, height, and materials of buildings. However, increased model complexity leads to increased cost, including higher computational power and more parameters requiring specifications. No model includes every possible specification for all exchange processes, so it’s in the interest of people constructing those LSMs to understand if there’s any performance improvement with increased complexity.
In this paper, Grimmond and his team analyzed 32 urban LSMs (ULSM) with a range of approaches. At first, all teams are only informed whether their sites are urban or not. Then, in four stages, Grimmond’s team released a controlled amount of information about the sites. Groups were told how their models performed but not how others performed. This experiment had three main objectives:
- To evaluate how ULSMs modeled urban energy balance fluxes in relation to varying degrees of info
- To evaluate and compare how models with similar characteristics and complexities performed relative to each other
- To use these findings to look for ways to improve future ULSMs
The conclusions were rather lengthy, and many related to specific variables defined in the methodology, so I’ll just mention a major one most pertinent to that outlined above. The results showed that simpler models often showed a net improvement with additional information but complex models did not, suggesting increased model complexity does not necessarily increase model performance. Additionally, the authors noted that only two modeling sites were compared in this research, and urges the need for future comparisons among a wider range of morphology and building site materials.
Thanks for reading,
Valerie Polukhtin
Comments:
All viewpoints are welcome but profane, threatening, disrespectful, or harassing comments will not be tolerated and are subject to moderation up to, and including, full deletion.