Human forecasts are based on a number of factors - forecasted weather (PFMs), numerical model (F1) forecasts, statistical model forecasts, narrative weather (discussions on local weather) and ozone transport. Due to the complexities in AQF, both numerical model and statistical model can fail. Human forecast can usually pick up the slacks not considered or mis-represented in the numerical models and statistical models.
The starting point of human forecast is to understand the weather with the aid of large-scale weather maps and local forecasts. Weather controls the ozone production, transport and destruction (aging). Given a certain weather pattern, there will be a corresponding ozone pattern. Wind, cloud, and precipitation are the most important factors to control ozone behaviour.
The next step is to look at the prevailing wind and the ozone coming from the upwind direction.
Because a single person may have bias, human forecasts are better to be brewed by a team with different expertise. Currently, IMAQS human forecast team comprises of Dr. Soontae Kim, Dr. Xiangshang Li and Dr. In-bo Oh. Dr. Kim has been a long time member of IMAQS and an expert in emissions. Dr. Li joined IMAQS in 2006. His specialty is statistics and weather patterns (climatology). Dr. Oh has been with IMAQS since Oct 2006. He specializes in meteorology.