Fire monitoring

1. A method comprising:

accessing, by a computer system, a database to obtain values for a plurality of features associated with a fire in a geographical region, the plurality of features comprising one or more satellite images at a first resolution, vegetation information for the geographical region, and weather data for the geographical region, each satellite image comprising a plurality of first cells associated with the geographical region, the first resolution defining a first size of each first cell;

generating a map of the geographical region, the map comprising a plurality of second cells having a second size, the second size being smaller than the first size, the map of the geographical region having a higher resolution than a resolution of the one or more satellite images;

estimating, using a machine-learning model; probability values for the second cells in the map based on the plurality of features, each probability value indicating if the second cell contains an active fire;

updating the map of the geographical region based on the probability values for the second cells, and

causing presentation of the map in a user interface.

2. The method of claim 1 , comprising:

before estimating the probability values, training the machine-learning model with values of the plurality of features from previous fires.

3. The method of claim 1 , wherein the plurality of features comprises land-related features that include land use, slope, aspect, elevation, and soil moisture.

4. The method of claim 1 , wherein the plurality of features comprises hour of the day, Normalized Difference Vegetation Index (NDVI), and Normalized Difference Built-up Index (NDBI).

5. The method of claim 1 , wherein the vegetation information comprises canopy height, canopy cover, canopy bulk density, and drought index; wherein the weather data comprises wind, precipitation, temperature, humidity, and cloud cover.

6. The method of claim 1 , comprising:

estimating, by the machine-learning model, second cells with embers that a ignite fires; and

presenting, in the map, an ember area with second cells that comprise embers that may ignite fires.

7. The method of claim 1 , comprising:

estimating a fire scar area with second cells that have already burnt; and

causing presentation in the user interface of the map with the fire scar area.

8. The method of claim 1 , wherein the one or more satellite images comprise images in a plurality of frequency bands.

9. The method of claim 1 , wherein the first size is a square with 375 meter sides and the second size is a square with 30 meter sides.

10. The method of claim 1 , comprising:

receiving observations about a state of the fire in the geographical region; and

updating the probability values for the second cells based on the received observations.

11. A system comprising:

a memory comprising instructions; and

one or more computer processors, wherein the instructions, when executed by the one or more computer processors, cause the one or more computer processors to perform operations comprising:

accessing a database to obtain values for a plurality of features associated with a fire in a geographical region, the plurality of features comprising one or more satellite images at a first resolution, vegetation information for the geographical region, and weather data for the geographical region, each satellite image comprising a plurality of first cells associated with the geographical region, the first resolution defining a first size of each first cell,

generating a map of the geographical region, the map comprising a plurality of second cells having a second size, the second size being smaller than the first size, the map of the geographical region having a higher resolution than a resolution of the one or more satellite images;

estimating, using a machine-learning model, probability values for the second cells in the map based on the plurality of features, each probability value indicating if the second cell contains an active fire;

updating the map of the geographical region based on the probability values for the second cells; and

causing presentation of the map in a user interface.

12. The system of claim 11 , wherein the instructions cause the one or more computer processors to perform operations comprising:

before estimating the probability values, training the machine-learning model, with values of the plurality of features from previous fires.

13. The system of claim 11 , wherein the plurality of features comprises land-related features that include land use, slope, aspect, elevation, and soil moisture; wherein the plurality of features comprises hour of the day, Normalized Difference Vegetation Index (NDVI), and Normalized Difference Built-up Index (NDBI).

14. The system of claim 11 , wherein the vegetation information comprises canopy height, canopy cover, canopy bulk density, and drought index; wherein the weather data comprises wind, precipitation, temperature, humidity, and cloud cover.

15. The system of claim 11 , wherein the instructions cause the one or more computer processors to perform operations comprising:

estimating; by the machine-learning model; second cells with embers that may ignite fires; and

presenting, in the map; an ember area with second cells with embers that may ignite fires.

16. A non-transitory machine-readable storage medium including instructions that, when executed by a machine, cause the machine to perform operations comprising:

accessing; by a computer system, a database to obtain values for a plurality of features associated with a fire in a geographical region, the plurality of features comprising one or more satellite images at a first resolution, vegetation information for the geographical region, and weather data for the geographical region, each satellite image comprising a plurality of first cells associated with the geographical region, the first resolution defining a first size of each first cell;

generating a map of the geographical region, the map comprising a plurality of second cells having a second size, the second size being smaller than the first size, the map of the geographical region having a higher resolution than a resolution of the one or more satellite images;

estimating, using a machine-learning model; probability values for the second cells in the map based on the plurality of features, each probability value indicating if the second cell contains an active fire;

updating the map of the geographical region based on the probability values for the second cells; and

causing presentation of the map in a user interface.

17. The non-transitory machine-readable storage medium of claim 16 , wherein the machine performs operations comprising:

before estimating the probability values, training the machine-learning model, with values of the plurality of features from previous fires.

18. The non-transitory machine-readable storage medium of claim 16 , wherein the plurality of features comprises land-related features that include land use, slope, aspect, elevation, and soil moisture; wherein the plurality of features comprises hour of the day, Normalized Difference Vegetation Index (NDVI), and Normalized Difference Built-up Index (NDBI).

19. The non-transitory machine-readable storage medium of claim 16 , wherein the vegetation information comprises canopy height, canopy cover, canopy bulk density, and drought index; wherein the weather data comprises wind, precipitation, temperature, humidity, and cloud cover.

20. The non-transitory machine-readable storage medium of claim 16 , wherein the machine performs operations comprising:

estimating, by the machine-learning model, second cells with embers that may ignite fires; and

presenting, in the map, an ember area with second cells with embers that may ignite fires.