Process for sensing-based crop nutrient management using limiting-rate estimation and yield response prediction
A sensing-based process for managing nutrient fertilization in crops. This invention uses reference plots in a field, fertilized at different pre-plant rates, including at least one plot fertilized at a lower rate than the majority of the field. Measurements from the plots are used to estimate the Current Limiting Rate (CLR), which is the lowest non-limiting nutrient rate at the time of sensing. Preferred embodiments are presented specifically for managing nitrogen in rain-fed and irrigated corn. These embodiments use the CLR, along with a model for CLR movement with time to determine nutrient rates for mid-season treatments that achieve specified yield-based or profit-based objectives. This invention's novel use of low-rate reference plots to estimate the CLR enable the detection of stress before the main area of a field is impacted, allowing for advance planning and intervention without sacrificing yield potential.
1. A method for applying a mid-season fertilizer application to a managed area within a field containing a crop, comprising:
a) creating reference plots with different, known, pre-plant nutrient rates, including at least 1 plot fertilized at a lower rate than a majority of the managed area;
b) estimating yield at a specified, low nutrient rate;
c) estimating yield at a high, non-limiting nutrient rate;
d) measuring a sensor-based reading of crop health or vigor in the reference plots created in (a) and elsewhere in the managed area using a handheld crop health sensor or chlorophyll meter, aircraft-mounted d standard, multi-spectral, or hyper-spectral camera, satellite-based optical imaging device, or a vegetation index derived from any of these sources, or other sensor capable of measuring crop health or vigor;
e) using the measurements from (d) to estimate a Current Limiting Rate (CLR) for the nutrient;
f) using the CLR estimate from (e) and a model for movement of the CLR with time for the crop to predict a Yield Response To Nutrient (YRTN) function;
g) using the predicted YRTN function from (f) to estimate a nutrient rate that achieves a yield-related or profit-related objective;
h) using the nutrient rate computed in (g) to design one or more mid-season nutrient treatments;
i) setting a variable-rate sprayer or other variable-rate fertilizer application equipment to execute the one or more mid-season nutrient treatments in the managed area; and
j) causing the variable-rate sprayer or other variable-rate fertilizer application equipment to execute the one or more mid-season nutrient treatments in the managed area.
2. The method of claim 1 wherein step (b) uses a yield measurement from a 0-rate or low-rate test plot in the field or a nearby field in a previous year or an average of such measurements over multiple years as the estimate of yield at a specified, low nutrient rate.
3. The method of claim 1 wherein step (b) includes:
i. looking up an estimate of soil organic matter (OM) content for the managed area in a soil database or measuring soil OM content with soil testing; and
ii. using the soil OM content from (i) to estimate the yield for the specified, low nutrient rate.
4. The method of claim 1 wherein step (c) uses an average yield for the field in a previous growing season, or an average or maximum of average yields for multiple previous growing seasons as the estimate of yield at a high, non-limiting rate nutrient rate.
5. The method of claim 1 wherein step (c) uses Yield Response To Nutrient curves from a database of nutrient trials to lookup the yield corresponding to a high, non-limiting nutrient rate.
6. The method of claim 1 wherein step (e) uses a quadratic plateau function as a model for an Index Response To Nutrient (IRTN) function, fits the IRTN function by minimizing l 1 error, Mean Square Error, or Weighted Mean Square Error on the measurements from step (d), and computes the CLR as the plateau of the IRTN function.
7. The method of claim 1 wherein step (f) uses a quadratic plateau function or an initial function from a database of nutrient trials, along with the estimates from (b) and (c) to define an initial estimate of the YRTN function, and adjusts this initial estimate to compute the predicted YRTN function.
8. The method of claim 7 wherein the model for the movement of the CLR with time is used to predict an input at which the YRTN function plateaus and wherein the adjustment made to the initial YRTN function is shifting the domain, scaling the domain, or a combination of shifting and scaling the domain of the initial estimate of the YRTN function so that the YRTN function plateaus at this input.
9. The method of claim 8 wherein the model for the movement of the CLR with time is a linear function in which the CLR is expected to increase a certain amount per day or per growing degree day.
10. The method of claim 1 wherein step (g) the nutrient rate is computed to achieve maximum profit, or is computed as the lowest nutrient rate that will maximize yield.
11. The method of claim 1 where in step (h) one or more, uniform, mid-season treatment(s) is/are computed so that the season-total applied nutrient rate for the managed area, excluding the reference plots, equals the rate computed in step (g).
12. The method of claim 1 wherein step (h) includes:
i. measuring the sensor-based reading of crop health or vigor in each location in the managed area;
ii. computing an effective nutrient rate for each location by inverting a function fit to the measurements made in (d); and
iii. computing the mid-season fertilizer application rate by subtracting the effective nutrient rate from (ii) from the target nutrient rate from (g) at each location.
13. A method for periodically applying a mid-season fertilizer application to an irrigation zone in a field containing a crop, comprising:
a) creating reference plots with different, known, pre-plant nutrient rates, including at least 1 plot fertilized at a lower rate than a majority of the irrigation zone;
b) measuring a sensor-based reading of crop health or vigor in the reference plots created in (a) and elsewhere in the irrigation zone using a handheld crop health sensor or chlorophyll meter, aircraft-mounted standard, multi-spectral, or hyper-spectral camera, satellite-based optical imaging device, or a vegetation index derived from any of these sources, or other sensor capable of measuring crop health or vigor;
c) using the measurements from (b) to estimate a Current Limiting Rate (CLR) for the nutrient;
d) using the CLR estimate from (c) and a model for movement of the CLR with time for the crop to compute a fertigation rate that will keep the field margin, defined as the cumulative applied nutrient rate for the area outside the reference plots minus the CLR, above a chosen value for a predetermined amount of time, for instance the time between planned sensing events;
e) causing fertilizer doses to be added to an irrigation system to apply fertilizer in the irrigation zone according to the fertigation rate;
f) causing the irrigation system to apply fertilizer in the irrigation zone according to the fertigation rate; and
f) repeating steps (b), (c), (d), (e), and (f) multiple times until the crop reaches maturity.
14. The method of claim 13 wherein step (c) uses a quadratic plateau function as a model for an Index Response To Nutrient (IRTN) function, fits the IRTN function by minimizing l 1 error, Mean Square Error, or Weighted Mean Square Error on the measurements from step (b), and computes the CLR as the plateau of the IRTN function.
15. The method of claim 13 wherein step (d) computes the fertigation rate to achieve the objective that the field Margin approximately equals O as the crop reaches maturity.
16. The method of claim 13 wherein step (d) computes the fertigation rate to achieve the objective that when the crop reaches maturity, the field Margin is negative by an amount computed to approximately maximize profit in the majority of the irrigation zone.