Generating digital models of relative yield of a crop based on nitrate values in the soil

1. A computer implement method comprising:

receiving, over a network at a server, a nitrogen recommendation request for a field, the field having multiple locations in the field, the request including one or more nitrate measurements in soil for each of the locations;

computing, by the server, a relative yield for a particular location in the field, for a crop, based, at least in part, on one or more nitrate measurements in soil for the particular location during a V6-V8 growth stage of the crop and on a digital model of relative crop yield;

computing, by the server, an optimal amount of nitrogen to apply to the field based, at least in part, on the relative yield for the particular location, by:

generating a model of crop yield based on one or more factors;

estimating the model of crop yield using a surrogate function based, at least in part, on one or more observations sampled from the model of crop yield, wherein the one or more observations belong to a corpus of observations;

determining one or more data points at which to sample the model of crop yield based on the surrogate function;

for each data point of the one or more data points, evaluating the data point using the model of crop yield to add a corresponding observation to the corpus of observations; and

determining the optimal amount of nitrogen to apply based on the corpus of observations; and

generating, by the server, one or more scripts based on, at least in part, the optimal amount of nitrogen; and

transmitting, to one or more agricultural implements and/or to one or more agricultural machines associated with the one or more agricultural implements, the one or more scripts to control an operating parameter of the one or more agricultural implements and/or the one or more agricultural machines in performing agricultural tasks in the particular location of the field consistent with the relative yield for the particular location and based on the optimal amount of nitrogen to apply to the field.

2. The method of claim 1 , wherein the digital model of relative crop yield is a quadratic-plateau function.

3. The method of claim 1 , further comprising sending, over the network to a client computer device, a nitrogen application recommendation, in response to the nitrogen recommendation request, for the field based, at least in part, on the computed optimal amount of nitrogen; and

wherein transmitting the one or more scripts is based on an authorization from the client computer device for the nitrogen application recommendation.

4. The method of claim 1 , further comprising, prior to receiving the request:

receiving, over a network at the server, first electronic digital data comprising a plurality of values representing, for each location of the multiple locations in the field, nitrate measurements in soil for the location during a V6-V8 growth stage of a crop;

receiving, over the network at the server, second electronic digital data comprising a plurality of values representing, for each location of the multiple locations, a crop yield corresponding to the received nitrate measurements in soil for the location;

determining, by the server, for each location of the multiple locations, based, at least in part, on a planting date of the crop, a planting density, a seed type, and the crop yield corresponding to the received nitrate measurements in soil for the location, a maximum yield of the crop with sufficient nitrate;

computing, by the server, for each location of the multiple of locations, a relative yield value comprising the crop yield corresponding to the received nitrate measurements in soil for the location divided by the determined maximum yield of the crop for the location; and

generating, by the server, the digital model of relative crop yield as a function of nitrate in soil during the V6-V8 growth stage of the crop based, at least in part, on the relative yield values and the nitrate measurements in soil.

5. The method of claim 4 , wherein determining, for each location of the multiple locations, a maximum yield of the crop with sufficient nitrate comprises computing a model of expected maximum yield for each location as a latent function based, at least in part, on the crop yield corresponding to the nitrate measurements in soil for the location; and

wherein the method further comprising sampling the model of expected maximum yield for each location.

6. The method of claim 4 , wherein the first electronic digital data comprises the plurality of values for each year of a plurality of years;

wherein the second electronic digital data comprises the plurality of values for each year of the plurality of yields;

wherein determining the maximum yield of the crop with sufficient nitrate includes determining the maximum yield of the crop with sufficient nitrate based, at least in part, on the received nitrate measurements in soil and the crop yields corresponding to the received nitrate measurements in soil for each location of the multiple locations and each year of the plurality of years; and

wherein computing a relative yield value for each location of the multiple locations includes computing a relative yield value for each location of the multiple locations for each year of the plurality of years.

7. The method of claim 1 , wherein the model of crop yield includes an objective function expressed as:

R ( F,D )= p Y EY ( F,D )− p F F−p I I F>0 .

8. One or more non-transitory computer-readable media, storing instructions which, when executed by one or more processors, cause performance of a method comprising the steps of:

receiving a nitrogen recommendation request for a field;

computing an optimal amount of nitrogen to apply to the field based, at least in part, on a relative yield for a particular location of the field, by:

generating a model of crop yield based on one or more factors;

estimating the model of crop yield using a surrogate function based, at least in part, on one or more observations sampled from the model of crop yield, wherein the one or more observations belong to a corpus of observations;

determining one or more data points at which to sample the model of crop yield based on the surrogate function;

for each data point of the one or more data points, evaluating the data point using the model of crop yield to add a corresponding observation to the corpus of observations; and

determining the optimal amount of nitrogen to apply based on the corpus of observations; and

generating one or more scripts based on, at least in part, the computed optimal amount of nitrogen; and

transmitting, to at last one agricultural machine, the one or more scripts to control an operating parameter of the at least one agricultural machine in performing agricultural tasks in the field based on the computed optimal amount of nitrogen.

9. The one or more non-transitory computer-readable media of claim 8 , wherein the instructions, when executed by the one or more processors, cause performance of:

in response to the nitrogen recommendation request, sending, to a client computer device, a nitrogen application recommendation for the field based, at least in part, on the computed optimal amount of nitrogen; and

transmitting the one or more scripts in response to an authorization from the client computer device for the nitrogen application recommendation.

10. The one or more non-transitory computer-readable media of claim 8 , wherein the instructions, when executed by the one or more processors, further cause performance of:

receiving first electronic digital data comprising a plurality of values representing, for each location of the multiple locations in the field, nitrate measurements in soil for the location during a V6-V8 growth stage of a crop;

receiving second electronic digital data comprising a plurality of values representing, for each location of the multiple locations, a crop yield corresponding to the received nitrate measurements in soil for the location;

determining, based, at least in part, on a planting date of the crop, a planting density, a seed type, and the crop yield corresponding to the received nitrate measurements in soil for the location, a maximum yield of the crop with sufficient nitrate;

computing, for each location of the multiple of locations, a relative yield value comprising the crop yield corresponding to the received nitrate measurements in soil for the location divided by the determined maximum yield of the crop for the location; and;

generating a digital model of relative crop yield as a function of nitrate in soil during the V6-V8 growth stage of the crop based, at least in part, on the relative yield values and the nitrate measurements in soil.

11. The one or more non-transitory computer-readable media of claim 10 , wherein the digital model of relative crop yield is a quadratic-plateau function.

12. The one or more non-transitory computer-readable media of claim 10 , wherein the instructions, when executed by the one or more processors, further cause performance of:

determining, for each location of the multiple locations, a maximum yield of the crop with sufficient nitrate by computing a model of expected maximum yield for each location as a latent function based, at least in part, on the crop yield corresponding to the nitrate measurements in soil for the location; and

sampling the model of expected maximum yield for each location.

13. The one or more non-transitory computer-readable media of claim 10 , wherein the first electronic digital data comprises the plurality of values for each year of a plurality of years;

wherein the second electronic digital data comprises the plurality of values for each year of the plurality of yields;

wherein determining the maximum yield of the crop with sufficient nitrate includes determining the maximum yield of the crop with sufficient nitrate based, at least in part, on the received nitrate measurements in soil and the crop yields corresponding to the received nitrate measurements in soil for each location of the multiple locations and each year of the plurality of years; and

wherein computing a relative yield value for each location of the multiple locations includes computing a relative yield value for each location of the multiple locations for each year of the plurality of years.