SNAP-CII is an algorithm that estimates aerosol levels from image data captured with everyday cameras, such as those in smartphones. The algorithm analyzes how sunlight is scattered by aerosols in the image data and classifies aerosol levels into three categories: low, medium, and high (see figure). Image data records the intensity of blue, green, and red light. When these wavelengths interact with aerosols and air molecules, they are scattered differently.

When aerosol levels are low, air molecules scatter more blue light, so the sky appears blue.
When aerosol levels are high, aerosols scatter blue, green, and red light at nearly the same intensity, so the sky appears whitish or grayish.
By training a machine-learning model on these differences, SNAP-CII can estimate aerosol levels with high accuracy.

Air quality — how clean the air is — may not be something you think about in everyday life.
However, you may inhale harmful substances without realizing it, and higher exposure can have adverse effects on one's health. A representative example is PM2.5, a type of aerosol.

Until now, information on the day's air quality has mainly been available from specialized websites or government agencies. But with SNAP-CII, you will be able to assess air quality yourself using a smartphone camera, etc.

Moreover, aerosols are a critical factor in predicting future climate change. Observing aerosols is therefore also extremely important for action concerning climate.

Currently, ten organizations are cooperating with us in pilot trials of SNAP-CII.（Coming Soon） Based on the feedback we receive, we will improve the system and ultimately aim to make it available for many people to use. By collecting observation data captured by users, we will build a scientifically valuable database and use it to support air pollution control, climate-change measures, and health protection.

Aerosol Observation Network with Citizen Participation

SNAP-CII research started in 2020 as a collaborative project between NICT, Kyushu Institute of Technology, Fukuoka University, and Tokyo Metropolitan University, with Weathernews Inc. joining in 2021.

In addition, we have received the cooperation of many partners — including the Kitakyushu Foundation for the Advancement of Industry, Science and Technology (FAIS), the University of Kitakyushu, Nagasaki University, and Meiji Gakuen Junior & Senior High School — for the installation of observation cameras and the provision of photographic data.

NEWS: