TOP > Atmospheric Environment and Applied Meteorology Research Department > First Laboratory

Summary Account

1. JRAero (Japanese Reanalysis for aerosol)


The Japanese Reanalysis for Aerosol (JRAero) is a global aerosol reanalysis product construnted by the Meteorological Research Institute (MRI) of the Japan Meteorological Agency (JMA) and the Research Institute for Applied Mechanics (RIAM) of Kyushu University (KU).

Available Data

Period 1 Jan. 2011 - 31 Dec. 2015
Temporal Resolution 6 hour
Horizontal Resoultion approximately 1.1 deg x 1.1 deg (TL159: 320 x 160 grid points)
Vertical Resolution 48 Layers Surface 0.01 hPa)
2-D variables AOD (total, dust, sea salt, BC, OC, sulfate), surface PM2.5 and PM10 [μ/m3], and deposition (dust) [g/m2/sec].
3-D variables extinction coefficient (total, dust, sea salt, BC, OC, sulfate), mixing ratio (dust, sea salt, BC, OC, sulfate) [kg/kg-air].

Data Access

Use of the JRAero is provided subjected to the terms and conditions.
Contact the JRAero developers via e-mail (yumimoto AT with "[JRAero]" in mail subject.

Term and Conditions

JRAero is a property of MRI/JMA and RIAM/KU. Use of the data is limited to non-profit purposes such as research and education. Commercial use is strictly prohibited under any circumstances. Redistribute the data to the third parties is prohibited. JRAero is provided AS IS, without warranties of any kind. MRI/JMA and RIAM/KU can not take any responsibilities that may be caused by the use of JRAero.

Acknowledgement and Feedback

 For the dataset source, please cite:

Yumimoto, K., Tanaka, T. Y., Oshima, N., and Maki, T.: JRAero: the Japanese Reanalysis for Aerosol v1.0, Geosci. Model Dev., 10. 3225-3253, 2017, doi:10.5194/gmd-10-3225-2017.

It would be very appreciated if you report your findings and suggestions and send a copy of your research outcome to the JRAero developers.

2. Research on Stratospheric Ozone (the Ozone Layer)

Ozone is a gas that is naturally in the atmosphere. Most ozone (about 90%) exists in the stratosphere from about 10 km to 50 km; the stratosphere region with the highest ozone concentration is commonly known as the ozone layer. The remaining, about 10%, is found in the troposphere. Stratospheric ozone is considered to be beneficial because it can absorb a large part of ultraviolet-B (UV-B) radiation, which is harmful to humans and other life forms. Total-column ozone is defined as the columnar density of ozone in the Earth’s atmosphere and those values are often reported in Dobson unit denoted as "DU". One Dobson unit (or 1 milli atm-centimeter) refers to a layer of gas that would be 10 µm thick under standard temperature. For example, if all atmospheric ozone from surface to top of the atmosphere is brought down to the surface of the Earth at 1013 hPa and 0 °C, the resulting layer of pure ozone would have a thickness of only about 3 mm. The severe ozone depletion of Antarctic ozone layer, known as the "ozone hole", was first reported in the mid-1980s by a researcher of MRI. The area of the ozone hall is often defined as the geographical region contained within column ozone amounts less than 220 DU.

During recent-past of several decades, a large fraction of human activity’s emission of the halogen chlorine and bromine, which is often referred to as ozone-depleting substances (ODSs), reach the stratosphere and lead to ozone depletion because chlorine and bromine atoms react to destroy ozone. The severe depletion of the Antarctic ozone layer is known as the ozone hole. It occurs in late winter and early spring due to the special atmospheric conditions and the effective ozone destruction by reactive halogen gases, i.e., ODSs. The production and consumption of ozone-depleting substances are controlled by a 1987 international agreement, Montreal Protocol on Substances that Depletion the Ozone Layer. Production and consumption of ODSs will be phased out and the ozone layer is expected to recover around the middle of this century. However, polar ozone depletion continues to occur in both hemispheres, so it is important for us to monitor the change in the ozone layer carefully in future. One of our laboratory’s aims is to detect any change in the ozone layer and to comprehensively understand its mechanism by using numerical model simulations and analyzing observation data.

Useful links to stratospheric ozone web sites:

Time-latitude cross sections of the total column ozone in Dobson Unit (DU)
Figure 1: Time-latitude cross sections of the total column ozone in Dobson Unit (DU). (Left): ozone simulated by our laboratory’s model, Meteorological Research Institute – Chemical Climate Model in version 2 (MRI-CCM2). (Right): the total ozone mapping spectrometer/ Solar Backscatter Ultraviolet Radiometer (TOMS/SBUV) satellite observation.

Figure 1 shows the seasonal variability of total ozone and the severe depleted ozone in the Antarctic area (i.e., the ozone hole). By comparing the simulated ozone (left panel) with the observation (right panel), we can say that MRI-CCM2 has an ability to reproduce a seasonal change of ozone layer and the ozone hole very well. It is noted that the white area over the pole areas represents the latitudes in which no observations can be made because the SBUV instrument cannot make observations in the polar night region because they relies upon backscattered sun light.

Data assimilation is the process by which observations are incorporated into a numerical model of a real system. The production of observations being combined with the results from a numerical model is called an "analysis", which is considered as the best estimate of the current state of the system. Applications of data assimilation are important in weather forecasting as well as ozone layer prediction. Currently, our laboratory has developed a sophisticated data assimilation method, Ensemble Kalman Filter (EnKF) technique, which has an ability to produce a better analysis of chemical species and aerosols.

how the deviations of model results from independent observation decrease as we assimilate model with several combinations of ozone observations
Figure 2: Ozone vertical distributions of (left) ozone concentration, (middle) model bias from the observation, and (right) the root mean square error between observation and model, averaged over the latitude of 60°–90°S in August–September 2006. Profiles are ozone sonde observations (black thick lines), model free run (gray lines), total column ozone assimilation (red lines), vertical profile ozone assimilation (green lines), and total column plus vertical profile ozone assimilation, respectively.

Figure 2 shows how the deviations of model results from independent observation decrease as we assimilate model with several combinations of ozone observations (total column amount and vertical profile). By comparing assimilated model results with model free run, we find that model results are improved by incorporating data assimilation, and we can obtain more reliable model results by combining total column and vertical profile observations.

These our research outcomes will be applied to the improvement UV information operation of Japan Meteorological Agency (JMA) after validation.

3. Research on aerosols

The atmospheric aerosol is a complex and dynamic mixture of solid and liquid particles suspended in the air. It consists of particles ranging from a few nanometers to several hundred micrometers. Particles in the atmosphere arise from natural sources, such as dust storms, spray from the sea surface, volcanic eruptions, vegetation fires, and from anthropogenic activities, such as the combustion of the fossil fuel. Dust storms or the yellow sand are phenomena in which mineral dust aerosols whirled up in the atmosphere by strong wind in the desert area of the East Asia, are transported downstream and deposited in East Asia. Imminent phenomena of the aerosol, which may cause the air pollution and give the damage to human society and health such as aggravation and respiratory illness of the visibility, are combined pollution of spring yellow sand and PM2.5, often reported by news recently. Fine particulate matter (PM2.5) is an air pollutant and consists of many compounds, including nitrates, sulfates, organics, metals, and soil or dust particles. Though substantial amounts of PM2.5 aerosols are emitted in the domestic of Japan, recent years about half of PM2.5 observed in Japan are thought to be transported from the region of remarkable economic development in East Asia.

In addition, aerosols have an influence on a climate and the weather of the earth through processes of scattering or absorbing atmospheric radiation, of acting as a cloud condensation nucleus, and of changing the optical characteristic of the cloud. It is important for us to evaluate various influences of the aerosol on climate changes such as the global warming.

In our laboratory, we have developed a global numerical simulation model of the aerosol called Model of Aerosol Species IN the Global AtmospheRe (MASINGAR) and studied many investigations of the aerosol relevant to the climate change as well as the environmental problems.

Horizontal distribution of aerosol optical thickness simulated by MASINGAR, including sulfate (white), black and organic carbon (green), dust (orange), and sea salt (blue).
Figure 3: Horizontal distribution of aerosol optical thickness simulated by MASINGAR, including sulfate (white), black and organic carbon (green), dust (orange), and sea salt (blue).

Figure 3 illustrates a simulated aerosol distribution by MASINGAR. It is a component of the Earth System Model (MRI-ESM) by which the Meteorological Research Institute uses studies of climate change. MASINGAR can simulate natural aerosols (dust, sea salt, and sulfate) using the atmospheric conditions such as wind direction and speed, temperature, humidity, soil wetness and snow equivalent water that MRI-ESM calculates. In addition, MASINGAR can simulate anthropogenic aerosols (sulfate, black and organic carbon) by using emission inventories provided by other institutes. MASINGAR has been used for the operational yellow sand prediction of JMA since 2004.

Effects on combining observation and model
Figure 4: Horizontal distributions of AOT derived from Himawari-8 data and modeled AOT at 0700 UTC on 15 April 2015 and 0200 UTC on 16 April 2015. (a and d) AOT derived from Himawari-8 data, (b and e) modeled AOT by simulation without data assimilation, and (c and f) modeled AOT by DA simulation. Black areas in (a) and (d) are areas of no data due to clouds and snow cover (Yumimoto et al. 2016).

Like ozone, combination between observation and model is significantly useful. Our laboratory has developed aerosol data assimilation techniques using the ensemble Kalman Filter. Figure 4 shows an example of the data assimilation of aerosol using Himawari-8 meteorological satellite. The model free run can simulate the tranport of Asian dust around Japan, but the area of the dust is overestimated. The data assimilatied result shows that the area of Asian dust is consistent with that of independent satellite observation. These research results will be incorporated into the update of JMA information (Kosa prediction) after validations.

Useful links

4. Research on air quality

In recent years, photochemical oxidant warnings were often reported. Photochemical oxidant warnings were issued in industrial zones and in the vicinity of Japan's mega city several decades ago. Today, however, photochemical smog warnings have been announced even in places far away from the big cities and industrial areas. One of the reasons is why a large amount of precursor of oxidant are emitted in regions of recent economic development in East Asia where and transported to Japan.

The original MRI-CCM of version one had mainly targeted the ozone layer (i.e., stratosphere). Our laboratory extended the simulation targeted area of MRI-CCM2 to the whole atmosphere from the surface to the mesopause (about 80 km), by including detail chemical processes in the troposphere. This model has been used for photochemical oxidant information of JMA. Today, we develop a new regional chemistry transport model (NHM-Chem) to simulate oxidant concentrations with higher resolution.

Simulated surface ozone concentration simulated by the regional chemical transport model, NHM-Chem in units of ppbv on 19 June 2013
Figure 5: Simulated surface ozone concentration simulated by the regional chemical transport model, NHM-Chem in units of ppbv on 19 June 2013. (Upper panels): observation data (Soramame-kun, MOE); (bottom panels): NHM-Chem simulations.

Figure 5 show results of simulated surface ozone concentration around Japan by the regional model, NHM-Chem. NHM-Chem with higher horizontal resolution (20km) has an ability to simulate better front line positions and better high ozone concentration areas, especially along the coast of the Sea of Japan, than those of global model, MRI-CCM2.

These research results will be incorporated into the update of photochemical oxidant information of JMA (Air pollution weather information) after validations.


Job titleName
HeadTakashi MAKI
Senior ResearcherHiroaki NAOE
Senior ResearcherTsuyoshi Thomas SEKIYAMA
Senior ResearcherNaga OSHIMA
ResearcherKeiichi KONDO

last update : Nov. 5 2014
All Rights Reserved, Copyright © 2003, Meteorological Research Institute, Japan