Remote sensing evidence of decadal changes in major tropospheric ozone precursors over East Asia.
Yunsoo Choi, Ph.D. Assistant Professor of Earth and Atmospheric Sciences, firstname.lastname@example.org, http://www.uh.edu/nsm/earth-atmospheric/people/faculty/yunsoo-choi/
Article Full Text: http://onlinelibrary.wiley.com/doi/10.1002/2016JD025663/full
Air pollution in East Asia has always been an issue that has a negative impact on human health, the economy and the environment. As opposed to stratospheric ozone whose role is same as an umbrella blocking UV radiation, ground-level ozone levels is harmful to live species, causing illnesses such as asthma and also associated with crop damage. The major challenge in tackling ozone pollution is the fact that it is formed by chemical reactions in the atmosphere, and not directly emitted into the air. Therefore we need to understand how its precursors including NO2 (nitrogen dioxide) and HCHO (formaldehyde) have evolved to understand how ozone concentrations have changed over time.
Due to the lack of sufficient in-situ observations of the mentioned gases, we analyzed the long-term trends of ozone precursors globally from observations provided by remote sensing satellites. The biggest benefit of using satellite data is their high spatial coverage that makes it possible to dig into ozone precursors changes over a large spatial domain. On the other hand, there appears a large number of pre-processing tasks to extract the qualified observations based on certain radiative or geometric flags. Performing these tasks is computationally burdensome, thereby the majority of former studies sought to resample the observations over larger grid cells. Using coarse resolution out some useful information about spatial patterns, essentially “throwing out the baby with the bath water.” Fortunately, Opuntia enabled us to carry out this step using the native spatial resolution of our data. We were able to read more than 13000 files from satellite, each containing more than 60×2000 observations (total 1.56 billion). Therefore, we were able to provide a high spatial resolution map of NO2 trends over East Asia. We found downward trends of NO2 over Japan, Taiwan, Seoul, more developed Chinese cities such as Guangzhou and Beijing, and upward trends in the Incheon Free Economic Zone in South Korea and the majority of northern regions of China. We observed recent strict regulations in North China had reduced levels of tropospheric NO2 mainly by the installation of selective catalytic reduction for power plants.
The observed NO2 trends were seen non-uniform spatially and temporally, so we offered a new perspective into data analysis called “trend classiﬁcation” to classify each trend into steady, positive-to-negative and negative-to-positive regions. From this novel perspective, we showed that the recent Chinese NO2 reductions occurred in predominantly in 2011 and 2012. Interestingly, following the decrease in Chinese NO2 “exports” in recent years, the contributions of local emissions in South Korea were also revealed. Using an inverse modeling approach, we linked these recent changes of NO2 with those of NOx emissions. To perform this task, a comprehensive weather and chemistry model was required to find the relationship between the satellite observations and the corresponding emissions. Here, we made use of Opuntia to conduct several extensive simulations in summertime 2010 and 2014. Our model output covered East Asia with a grid resolution of 15 km (269 × 249) with 15 vertical layers in June through August 2010 and 2014. Therefore we had more than 2 billion simulated values for each gas/particle species.
In addition to NO2, HCHO contributes to ozone formation through a complex non-linear chemical process. We found increasing trends of HCHO over the majority of East Asia excluding the southern part of China. In contrast to NO2, whose major emissions are anthropogenic, HCHO is produced directly and indirectly from both biogenic and anthropogenic sources. Accordingly, we investigated long-term trends of several driving factors for biogenic emissions using a biogenic emission model for a ten years period. Certainly, conducting this long-term simulation demands huge space and computational resources, all of which were nicely supported by Opuntia. More than 10 T storage was provided by CACDS to run the model. We found the biogenic emissions did not change much in the southern China, thereby the reduction of HCHO was primarily a result of anthropogenic emissions reductions.
Figure caption. Left panel: statistically significant trends of OMI tropospheric NO2 from 2005 to 2014. Numbers indicate the following cities: (1) Guangzhou, (2) Shanghai, (3) Beijing, (4) Shantou, (5) Shenzhen, (6) Seoul, (7) Incheon, (8) Busan, (9) Daegu, (10) Daejeon, (11) Tokyo, (12) Yokohama, (13) Osaka, and (14) Nagoya. Black and orange stars denote the Bohai and Yellow Seas respectively.