Page 639 - NGTU_paper_withoutVideo

Modern Geomatics Technologies and Applications

PM 2.5 concentrations in the nearest Inversion intensity parameter Meteorological data of synoptic stations
two neighbourhoods

Generate a raster map for each
parameter by Kriging method

Determine the parameters at each air
pollution measuring point
Pre-processing of meteorological data
and the concentration of pollutants
Filling Gap

Smoothing Data
Time Data

Creating training datasets based on Table 3

Decision Tree algorithm Topographic
Data
C4.5 Algorithm CART algorithm

Cross-validation, determining the Overall Accuracy and Kappa
index

Generate a pollution classification map
and forecast و یجنسرابتعا

Fig.1. Proposed method

The values of inversion intensity and interpretation of these values are shown in Table 1[14]:

Table 1 Inversion intensity values and interpretation of these values
Stability Division Range of temperature changes in
classification classification the vertical direction (C / 100m◦)

Very unstable A <-1.9
∆
∆
Unstable medium B -1.9<= <-1.7
∆
∆
Slightly unstable C -1.7<= <-1.5
∆
∆
Neutral D -1.5<= <-0.5
∆
∆
A little stable E -0.5<= <1.5
∆
∆
Stable medium F
∆
1.5<= <0.4
∆
Very stable G ∆
0.4<=
∆

3