SYSTEM, METHOD AND PROGRAM PRODUCT FOR PROVIDING POPULACE CENTRIC WEATHER FORECASTS

• US 20130085673A1
• Filed: 10/03/2011
• Published: 04/04/2013
• Est. Priority Date: 10/03/2011
• Status: Active Application

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Netto, Marco Aurelio Stelmar, Herrmann, Ricardo Guimaraes, Mantripragada, Kiran, Souza, Cleidson Ronald Botelho De, Cavalcante, Victor Fernandes, Real, Lucas Correia Villa
• Granted Patent: US 10,267,950 B2
• Time in Patent Office: Days
• US Class Current: 702/3
• CPC Class Codes: G01W 1/10 Devices for predicting weather conditions computers per se G06; display devices G09

Abstract

A populace centric weather forecast system, method of forecasting weather and a computer program product therefor. A forecasting computer applies a grid to a forecast area and provides a weather forecast for each grid cell. Area activity data sources indicate human activity in the forecast area. A dynamic selection module iteratively identifies grid cells for refinement in response to the weather forecast and to indicated/expected human activity. The dynamic selection module provides the forecasting computer with a refined grid for each identified grid cell in each iteration. The forecasting computer provides a refined weather forecast in each iteration.

Claims

1. A populace centric weather forecast system comprising:

a forecasting computer applying a grid to a forecast area and providing a weather forecast for each grid cell;

one or more area activity data sources indicating human activity in said forecast area; and

a dynamic selection module iteratively identifying grid cells for refinement responsive to said weather forecast and indicated said human activity, said dynamic selection module providing said forecasting computer with a refined grid for each identified said grid cell in each iteration, said forecasting computer providing a more comprehensive weather forecast in said each iteration.

2. A populace centric weather forecast system as in claim 1, wherein said one or more area activity data sources include historical data, real time data and projected data.

3. A populace centric weather forecast system as in claim 2, wherein said one or more area activity data sources include event data, demographic data, historical incident data, weather sensor data and geodata.

4. A populace centric weather forecast system as in claim 3, wherein said demographic data includes real time demographic data and said weather sensor data includes real time sensor data.

5. A populace centric weather forecast system as in claim 1, wherein in each iteration said dynamic selection module identifies as complete any cells in the most recent iteration with weather inactivity or human inactivity during the forecast period.

6. A populace centric weather forecast system as in claim 5, wherein said dynamic selection module ignores cell areas previously identified complete in providing refined grids.

7. A populace centric weather forecast system as in claim 6, wherein in each said iteration said forecasting computer overlays a current grid on an area map from a geodatabase and forecasts weather for each grid cell responsive to real time weather sensor data.

8. A method of forecasting weather, said method comprising:

overlaying an initial grid on a forecast area, said initial grid segmenting said forecast area into cells;

forecasting the weather during a forecast period for the area enclosed in each cell;

marking complete each cell with weather or human inactivity;

determining if all cells are marked complete;

providing a refined grid for every cell not marked complete;

overlaying said refined grid on a respective forecast area not marked complete; and

returning to forecasting the weather, wherein in the next iteration the weather is forecast for each area enclosed in each refined grid cell.

9. A method of forecasting weather as in claim 8, wherein when all cells are marked complete, the weather forecast for said forecast area is complete.

10. A method of forecasting weather as in claim 8, wherein marking cells complete comprises:

receiving activity data for each cell from activity data sources; and

determining for said each cell whether the current forecast and said activity data indicate weather inactivity or human inactivity for the respective cell area during said forecast period.

11. A method of forecasting weather as in claim 10, wherein said activity data includes historical data, real time data and projected data.

12. A method of forecasting weather as in claim 10, wherein said activity data includes event data, demographic data, historical incident data, weather sensor data and geodata.

13. A method of forecasting weather as in claim 8, wherein cells marked complete include uninhabited cells with inclement weather forecast and heavily populated cells with clear weather forecast.

14. A method of forecasting weather as in claim 8, wherein overlaying said initial grid on said forecast area comprises overlaying said initial grid on a map of said forecast area having a first resolution, said refined grid being overlaid on a higher resolution map of said respective cell area.

15. A method of forecasting weather as in claim 14, wherein cells are further marked complete whenever a forecast has been provided for said respective forecast area with said refined grid overlaid on the highest resolution map.

20. A computer program product for forecasting weather as in claim 13, wherein marking cells comprises determining whether historical, real time and projected activity data sources indicate weather activity and human activity during said forecast period in said forecast area, said activity data sources including sources of event data, demographic data, historical incident data, forecast data, weather sensor data and geodata.

16. A computer program product for forecasting weather, said computer program product comprising a computer usable medium having computer readable program code stored thereon, said computer readable program code causing a computer executing said code to:

overlay an initial grid on a forecast area, said grid segmenting said forecast area into cells; and

iteratively

forecast the weather for the area enclosed in each cell, said forecast being for a forecast period;

mark as complete each cell with weather inactivity or human inactivity during said forecast period;

identify any cells not marked complete;

provide a refined grid for the area in every unmarked cell;

overlay said refined grid on a respective unmarked forecast area; and

return to forecast the weather in a next iteration, wherein the weather is forecast for each area enclosed in each refined grid cell.

17. A computer program product for forecasting weather as in claim 16, wherein overlaying said initial grid on said forecast area comprises retrieving a map of said forecast area having a first resolution and overlaying said initial grid on said map.

18. A computer program product for forecasting weather as in claim 17, wherein overlaying said refined grid comprises retrieving a higher resolution map of said respective forecast area, said refined grid being overlaid on said higher resolution map, and wherein said cells in said refined grid are further marked complete whenever a forecast has been provided for the highest resolution map for said respective forecast area.

19. A computer program product for forecasting weather as in claim 16, wherein conditions in said forecast area are modeled with a fluid dynamics model to forecast the weather and whenever all cells are marked complete, the weather forecast for said forecast area is complete for said forecast period.

21. A computer program product for forecasting weather in a populace centric weather forecast, said computer program product comprising a computer usable medium having computer readable program code stored thereon, said computer readable program code comprising:

computer readable program code means for applying a grid to a forecast area;

computer readable program code means for providing a weather forecast for each grid cell in said forecast area;

computer readable program code means for indicating weather and human activity in said forecast area;

computer readable program code means for identifying grid cells for refinement responsive to indicated said weather and human activity in each grid cell; and

computer readable program code means for providing a refined grid for said each identified said grid cell, said computer readable program code means for applying said grid applying said refined grid, a more comprehensive said weather forecast being provided for identified said grid cells in the next iteration, a populace centric weather forecast being iteratively provided.

22. A computer program product for forecasting weather in a populace centric weather forecast as in claim 21, wherein said computer readable program code means for indicating weather and human activity comprises computer readable program code means for storing event data, demographic data, historical incident data, forecast data and a geodatabase.

23. A computer program product for forecasting weather in a populace centric weather forecast as in claim 22, further comprising computer readable program code means for receiving real time demographic data and weather sensor data.

24. A computer program product for forecasting weather in a populace centric weather forecast as in claim 22, wherein said event data includes data on historical events and scheduled events.

25. A computer program product for forecasting weather in a populace centric weather forecast as in claim 22, wherein said computer readable program code means for indicating weather and human activity identifies as complete any cells in the most recent iteration with weather or human inactivity, previously identified complete cell areas being ignored in providing weather for each said refined grid and a current grid being overlaid on an area map from said geodatabase, weather being forecast for each grid cell responsive to real time sensor data.