Step 1: manage data confidence

A - Plan how to integrate confidence level aspect in the data and the processing objectives

• Understand confidence level. Three confidence levels exist for this type of data.
  • Low Confidence (0-29%),
  • Nominal Confidence (30-79%)
  • High Confidence (80-100%)
• Calculate the confidence level attribute in the fire point layer
  • Plan field name (conf_l)
• Count fire points in the administrative polygon layer for each confidence level
  • Plan field names (nb_high_fire, nb_nominal_fire, nb_low_fire)

B - Calculate the confidence level attribute in the fire point layer

• Add algorithms
  • Field calculator (rename it “Field calculator conf_l” to have a better management of your process)
• Calculate “conf_l” with an expression using “CONFIDENCE” attribute of the layer
  • Input layer “Fire point layer”
  • Field name: conf_l
  • Type: string
  • Length: 100
  • Expression:

    if("CONFIDENCE">30 and"CONFIDENCE"<80 ,'nominal',if("CONFIDENCE"<30,'low','high')) 
    

    

C - Count fire points in the administrative polygon layer for each confidence level

• Add algorithms
  • Use 3 time Extract by expression (rename it to have a better management of your process, one for high fires, one for nominal fires and one for low fire), ex: “Extract by expression high fires”
    • Select input layer (Using algorithm output): Calculated from algorithm “Field calculator conf_l”
    • Set expression: “conf_l” = ‘high’
    • Same for the other confidence level (nominal and low)

💡You can copy, paste the algorithm selected in the Graphical modeller.

• Use 3 time Count point in polygons (rename it to have a better management of your process, one for high fires, one for nominal fires and one for low fire), ex: “Count point in polygon high fire”
  • Modify the one we add in “Our first model” part
    • Input polygon layer: administrative layer
    • Input point layer: Matching features from algorithm “Extract by expression high fire”
    • Count field name: nb_high_fire
  • Do same for nominal fire
    • Input polygon layer: Count from algorithm “Count point in polygon high fire”
    • Input point layer: Matching features from algorithm “Extract by expression nominal fire”
    • Count field name: nb_nominal_fire
  • Do same for low fire
    • Input polygon layer: Count from algorithm “Count point in polygon nominal fire”
    • Input point layer: Matching features from algorithm “Extract by expression low fire”
    • Count field name: nb_low_fire

D - Manage algorithm dependencies to be sure they run in a proper order. Go to the first algorithm and click on the three dots of dependencies and select the algorithm parent.

• “Extract by expression high fire” depend “Field calculator conf_l”
• “Count point in polygon high fire” depend “Extract by expression high fire”
• “Extract by expression nominal fire” depend “Count point in polygon high fire”
• … until “Count point in polygon low fire”

E - Fill output name of “Field calculator conf_l” and “Count point in polygon low fire” algorithms

• Check if all calculated fields are ok

F - Save as you model, in case further developments break its execution. It will help you to get back on a solid base, if you mess up the processing.

Model output: Model_on_fire_v2.1.model3