Algorithm 1.

Global air pollution prediction with Gaussian air spread plume.

Algorithm predict_air_pollution (class pollution_area *current_pollution_area, class wind_info *wind)

input: current_pollution_area  // the properties of global pollution area such as max() and min().

wind  // the properties of a wind such as direction, speed.

output: predicted pollution level // the predicted value in 10, 30, 60 minutes

method:

// check the progress direction and get predicted pollution level

for each time // 10, 30 60 minutes

// get the moving position in each time

distance = wind.speed * time

for each position of current pollution area such as max(), min(), boundary

target.x = current_pollution_area.position.x + distance * cos(wind.direction * pi / 180)

target.y = current_pollution_area.position.y + distance * sin(wind.direction * pi / 180)

target.value = current_pollution_area.position.value

// pollution value prediction at each position in each time

pollution_level[time][position] = Gaussian_air_pollution_dispersion(time, current_pollution_area, target)

dangerous_rate[time][position]= pollution_level[time][position] / AQI(level_5) * 100 * gradient

endfor

endfor

return pollution_level[time][position], dangerous_rate[time][position]

end