Suggesting sensor placements for improving emission inventory

1. A computer-implemented method, comprising:

simulating the concentration of each of one or more pollutants over a given geographic area for a given temporal period, wherein the given geographic area is covered by a sensor network grid, and wherein said simulating comprises applying one or more pre-determined models to a first version of an emission inventory corresponding to the given geographic area, wherein the emission inventory comprises a pre-determined estimation of the amount of each of the one or more pollutants discharged into the given geographic area within the given temporal period;

computing any discrepancy between the simulated concentration of each of the one or more pollutants and historical data pertaining to the one or more pollutants;

applying multiple scaling vectors to the first version of the emission inventory to generate multiple scaled versions of the emission inventory, wherein each of the scaling vectors comprises a correction factor that is multiplied with the first version of the emission inventory to generate a distinct scaled version of the emission inventory;

for each of the scaled versions of the emission inventory, simulating the concentration of each of the one or more pollutants over the given geographic area for the given temporal period by applying the one or more pre-determined models to the scaled version of the emission inventory;

for each of the scaled versions of the emission inventory, computing any discrepancy between the simulated concentration of each of the one or more pollutants and the historical data;

identifying each of the scaling vectors that corresponds with a discrepancy that is within a pre-determined range;

applying the identified scaling vectors to an emission inventory value attributed to each of multiple grid blocks within the sensor network grid, to generate a set of scaled emission inventory values for each of the multiple grid blocks;

determining one or more of the grid blocks having greater scaled emission inventory value variability within relative to the other grid blocks, wherein greater scaled emission inventory value variability corresponds to less confidence in a respective set of scaled emission inventory values; and

outputting a sensor placement suggestion to a user, wherein the sensor placement suggestion identifies each of the one or more determined grid blocks as a candidate for placement of a sensor to increase emission inventory accuracy for the given geographic area;

wherein the method is carried out by at least one computing device.

2. The computer-implemented method of claim 1 , wherein the one or more pre-determined models comprise one or more weather-related models.

3. The computer-implemented method of claim 2 , wherein the one or more weather-related models comprises a temperature model.

4. The computer-implemented method of claim 2 , wherein the one or more weather-related models comprises a wind speed model.

5. The computer-implemented method of claim 1 , wherein the one or more pre-determined models comprise one or more chemistry transport models.

6. The computer-implemented method of claim 1 , wherein the one or more pre-determined models comprise one or more weather-related models and one or more chemistry transport models.

7. The computer-implemented method of claim 1 , comprising:

filtering the one or more determined grid blocks based on one or more constraints pertaining to sensor location within the given geographic area.

8. The computer-implemented method of claim 7 , wherein said one or more constraints comprises one or more no-construction zones within the given geographic area.

9. The computer-implemented method of claim 7 , wherein said one or more constraints comprises one or more pre-determined sensitive areas within the given geographic area.

10. The computer-implemented method of claim 1 , comprising:

generating a variability map for the sensor network grid based on the generated set of scaled emission inventory values attributed to each of the multiple grid blocks.

11. The computer-implemented method of claim 10 , comprising:

ranking the identified scaling vectors, in descending order of their variance, based on the generated variability map.

12. The computer-implemented method of claim 1 , wherein the sensor placement suggestion comprises a suggestion to add an additional sensor at each of the one or more determined grid blocks.

13. The computer-implemented method of claim 1 , wherein the sensor placement suggestion comprises a suggestion to reposition an existing sensor to each of the one or more determined grid blocks.

14. A computer program product comprising a computer readable storage medium having program instructions embodied therewith, the program instructions executable by a device to cause the device to:

simulate the concentration of each of one or more pollutants over a given geographic area for a given temporal period, wherein the given geographic area is covered by a sensor network grid, and wherein said simulating comprises applying one or more pre-determined models to a first version of an emission inventory corresponding to the given geographic area, wherein the emission inventory comprises a pre-determined estimation of the amount of each of the one or more pollutants discharged into the given geographic area within the given temporal period;

compute any discrepancy between the simulated concentration of each of the one or more pollutants and historical data pertaining to the one or more pollutants;

apply multiple scaling vectors to the first version of the emission inventory to generate multiple scaled versions of the emission inventory, wherein each of the scaling vectors comprises a correction factor that is multiplied with the first version of the emission inventory to generate a distinct scaled version of the emission inventory;

for each of the scaled versions of the emission inventory, simulate the concentration of each of the one or more pollutants over the given geographic area for the given temporal period by applying the one or more pre-determined models to the scaled version of the emission inventory;

for each of the scaled versions of the emission inventory, compute any discrepancy between the simulated concentration of each of the one or more pollutants and the historical data;

identify each of the scaling vectors that corresponds with a discrepancy that is within a pre-determined range;

apply the identified scaling vectors to an emission inventory value attributed to each of multiple grid blocks within the sensor network grid, to generate a set of scaled emission inventory values for each of the multiple grid blocks;

determine one or more of the grid blocks having greater scaled emission inventory value variability within relative to the other grid blocks, wherein greater scaled emission inventory value variability corresponds to less confidence in a respective set of scaled emission inventory values; and

output a sensor placement suggestion to a user, wherein the sensor placement suggestion identifies each of the one or more determined grid blocks as a candidate for placement of a sensor to increase emission inventory accuracy for the given geographic area.

15. The computer program product of claim 14 , wherein the one or more pre-determined models comprise one or more weather-related models and one or more chemistry transport models.

16. The computer program product of claim 14 , wherein the program instructions executable by a computing device further cause the computing device to:

filter the one or more determined grid blocks based on one or more constraints pertaining to sensor location within the given geographic area.

17. The computer program product of claim 14 , wherein the program instructions executable by a computing device further cause the computing device to:

generate a variability map for the sensor network grid based on the generated set of scaled emission inventory values attributed to each of the multiple grid blocks.

18. The computer program product of claim 17 , wherein the program instructions executable by a computing device further cause the computing device to:

rank the identified scaling vectors, in descending order of their variance, based on the generated variability map.

19. A system comprising:

a memory; and

at least one processor operably coupled to the memory and configured for to perform the steps of:

simulating the concentration of each of one or more pollutants over a given geographic area for a given temporal period, wherein the given geographic area is covered by a sensor network grid, and wherein said simulating comprises applying one or more pre-determined models to a first version of an emission inventory corresponding to the given geographic area, wherein the emission inventory comprises a pre-determined estimation of the amount of each of the one or more pollutants discharged into the given geographic area within the given temporal period;

computing any discrepancy between the simulated concentration of each of the one or more pollutants and historical data pertaining to the one or more pollutants;

applying multiple scaling vectors to the first version of the emission inventory to generate multiple scaled versions of the emission inventory, wherein each of the scaling vectors comprises a correction factor that is multiplied with the first version of the emission inventory to generate a distinct scaled version of the emission inventory;

for each of the scaled versions of the emission inventory, simulating the concentration of each of the one or more pollutants over the given geographic area for the given temporal period by applying the one or more pre-determined models to the scaled version of the emission inventory;

for each of the scaled versions of the emission inventory, computing any discrepancy between the simulated concentration of each of the one or more pollutants and the historical data;

identifying each of the scaling vectors that corresponds with a discrepancy that is within a pre-determined range;

applying the identified scaling vectors to an emission inventory value attributed to each of multiple grid blocks within the sensor network grid, to generate a set of scaled emission inventory values for each of the multiple grid blocks;

determining one or more of the grid blocks having greater scaled emission inventory value variability within relative to the other grid blocks, wherein greater scaled emission inventory value variability corresponds to less confidence in a respective set of scaled emission inventory values; and

outputting a sensor placement suggestion to a user, wherein the sensor placement suggestion identifies each of the one or more determined grid blocks as a candidate for placement of a sensor to increase emission inventory accuracy for the given geographic area.

20. A computer-implemented method, comprising:

simulating the concentration of each of one or more pollutants over a given geographic area for a given temporal period, wherein the given geographic area is covered by a sensor network grid, and wherein said simulating comprises applying a coupled set of pre-determined models to a first version of an emission inventory corresponding to the given geographic area, wherein the emission inventory comprises a pre-determined estimation of the amount of each of the one or more pollutants discharged into the given geographic area within the given temporal period, and wherein said applying the coupled set of pre-determined models comprises:

using a weather conditions model to estimate one or more weather conditions at one or more locations of the given geographic area during the given temporal period; and

using a chemical transport model to estimate the dispersion of each of the one or more pollutants within the given geographic area based on (i) the first version of the emission inventory and (ii) the estimated weather conditions;

computing any discrepancy between the simulated concentration of each of the one or more pollutants and historical data pertaining to the one or more pollutants;

applying multiple scaling vectors to the first version of the emission inventory to generate multiple scaled versions of the emission inventory, wherein each of the scaling vectors comprises a correction factor that is multiplied with the first version of the emission inventory to generate a distinct scaled version of the emission inventory;

for each of the scaled versions of the emission inventory, simulating the concentration of each of the one or more pollutants over the given geographic area for the given temporal period by applying the one or more pre-determined models to the scaled version of the emission inventory;

for each of the scaled versions of the emission inventory, computing any discrepancy between the simulated concentration of each of the one or more pollutants and the historical data;

identifying each of the scaling vectors that corresponds with a discrepancy that is within a pre-determined range;

applying the identified scaling vectors to an emission inventory value attributed to each of multiple grid blocks within the sensor network grid, to generate a set of scaled emission inventory values for each of the multiple grid blocks;

for each of the grid blocks, determining the variability within its set of scaled emission inventory values, wherein greater variability within a set of scaled emission inventory values corresponds to less confidence in the set of scaled emission inventory values;

ranking the grid blocks, in descending order, by amount of variability; and

outputting a sensor placement suggestion to a user, wherein the sensor placement suggestion identifies each of the one or more highest-ranked grid blocks as a candidate for placement of a sensor to increase emission inventory accuracy for the given geographic area;

wherein the method is carried out by at least one computing device.