Fig 1.
Three provinces highlighted in the Thailand map indicate the locations of Nakhon Si Thammarat (Pink), Krabi (Yellow), and Bangkok (Red).
The map in this figure was produced using ArcGIS version 10.4 (Esri, Redlands, CA, USA). Source of shapefile: United Nations Office for the Coordination of Humanitarian Affairs https://data.humdata.org/dataset/thailand-administrative-boundaries.
Fig 2.
Monthly dengue cases (bar), rainfall (blue dotted line), and temperature (orange line) between 2014 and 2018 in the study areas.
The left y-axis represents the number of dengue cases, the right y-axis represents the temperature (LST) and the amount of rainfall (RF).
Fig 3.
Eight outdoor container types detected from street view images [4], from left to right: Jar, bucket, discarded tire, potted plant, bin, bowl, miscellaneous short open, miscellaneous tall open.
Fig 4.
Container density by population (A) Bangkok (B) Nakhon Si Thammarat (C) Krabi province. White color represents sub-districts with no data. The choropleth map in this figure was produced using ArcGIS version 10.4 (Esri, Redlands, CA, USA). Source of shapefile: United Nations Office for the Coordination of Humanitarian Affairs https://data.humdata.org/dataset/thailand-administrative-boundaries.
Table 1.
List of candidate variables in the dataset.
Table 2.
List of variables for each province.
Table 3.
Models and model equations in standard mathematical equations.
Table 4.
Performance of the models on training data (2015–2017) and test data (2018).
Fig 5.
Scatter plots of predicted vs actual values of dengue incidence for LMER+C models for Bangkok, Nakhon Si Thammarat, and Krabi province for the training (2015–2017) and testing (2018) data.
The maximum p-value of all panels is less than 2.384x10-10. The solid line is a linear trend line which is an indication of the linear (Pearson) correlation between the two variables. (Note: shading shows the 99% confidence interval).
Fig 6.
The risk maps of Bangkok, Nakhon Si Thammarat, Krabi for the training (2015–2017), and test (2018) data.
White color represents sub-districts with no data. The choropleth map in this figure was produced using ArcGIS version 10.4 (Esri, Redlands, CA, USA). Source of shapefile: United Nations Office for the Coordination of Humanitarian Affairs https://data.humdata.org/dataset/thailand-administrative-boundaries.
Table 5.
The coefficients of the LMER+C model using training data (2015–2017) for Bangkok province.
Significant variables with a p-value <0.05 are indicated with boldface.
Table 6.
The coefficients of the LMER+C model using training data (2015–2017) for Nakhon Si Thammarat province.
Significant variables with a p-value <0.05 are indicated with boldface.
Table 7.
The coefficients of the LMER+C models using training data (2015–2017) for Krabi province.
Significant variables with a p-value <0.05 are indicated with boldface.
Fig 7.
The ranges of sensitivity index (S100) for each province.
Fig 8.
The proportion of decrease in dengue cases in (A) Bangkok, (B) Nakhon Si Thammarat, (C) Krabi province.