The Optimization of Visibility Monitoring Network in Guangdong

Visibility is an important indicator to measure the atmospheric transparency conditions, which not only reflects the regional air quality conditions, but also closely relates to human life. Under current visibility conditions, it is crucial to implement an extensive long term stations for visibility monitoring network to track changes in visibility and determine causal mechanism for the visibility impairment in the region. The visibility observation network in Guangdong includes artificial monitoring network and the visibility sensor network. Artificial visibility observation is carried out at 86 meteorological stations every day, instrument observation is carried out at 39 stations, and the spatial distribution of stations is intensive but quite uneven. Therefore, a new method is carried out aiming at the optimization the overall arrangement of visibility monitoring network.
    Based on the dataset (daily visibility and relative humidity from the 86 meteorological stations in Guangdong), the method can be applied to optimization of the overall arrangement of the establishing operational visibility sensor network in Guangdong, which will substitute the artificial monitoring network. The figure-of-merit (FOM) and the spheres of influence (SOI) are calculated, and the most desirable location is ranked and identified using the resultant FOM field. The spatial coverage for each of stations is determined by the SOI. The determination of the minimum number of stations required is carried out by deleting lower ranking stations if more than 50% of its effecting area is covered by higher ranking stations. Besides, taking the local terrain, background stations, and other factors into account, it's suggested 51 stations are required.
    Above results can be applied in establishing operational visibility sensor network in Guangdong, which will substitute the artificial monitoring network. The related methods are also applicable to the overall arrangement for monitoring network of other variables.