# US 2022/0018990 A1 · Generating and conveying comprehensive weather insights at fields for optimal agricultural decision making

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2022/0018990 A1
- Document ID: US20220018990A1
- Status: Published application
- Kind: A1
- Application number: 17/316,210
- Filed: 2021-05-10
- Published: 2022-01-20
- Earliest claimed: 2019-03-27
- Assignees: Climate LLC
- Applicants: The Climate Corporation
- Inventors: Frederick Anthony Eckel; Gregory Reid Herman; Christopher Goodman
- Web page: https://patentgrep.com/patent/US20220018990A1

## CPC

- G01W 1/10: Devices for predicting weather conditions (computers per se G06; display devices G09)
- G06N 5/02: Knowledge representation; Symbolic representation
- G06N 5/04: Inference or reasoning models
- G06N 20/00: Machine learning
- G01W 2203/00: Real-time site-specific personalized weather information, e.g. nowcasting
- G06N 3/126: Evolutionary algorithms, e.g. genetic algorithms or genetic programming
- Y02A 90/10: Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
- G01W: METEOROLOGY (radar, sonar, lidar or analogous systems, designed for meteorological use G01S13/95, G01S15/88, G01S17/95)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- Y02A: TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE

## Abstract

In an embodiment, a computer-implemented method of generating and displaying a comprehensive depiction of a weather element comprises: based on archived forecast model and observed data, training a machine learning model; calibrating current forecast data by applying the machine learning model to yield a calibrated forecast probability density function; displaying graphical representation of recently observed data and calibrated forecast probability density.

## Claims

### Claim at position 1

-20. (canceled)

### Claim 21

A computer-implemented method providing improvements in agricultural science by optimizing weather forecasting practices, the method comprising:
receiving historic weather data associated with a field, the historic weather data relating to weather conditions at the field over a historic period of time;
generating and digitally storing, based on the historic weather data, a weather model for predicting weather conditions at the field; receiving contemporary weather data for the field, the contemporary weather data corresponding to weather conditions at the field at a recent time;
modifying the weather model based on the contemporary weather data by generating, using the weather model, expected weather data for the field at the recent time;
determining an output difference between the expected weather data and the contemporary weather data;
adjusting the weather model to converge toward the contemporary weather data;
using the adjusted weather model, generating one or more weather forecasts;
causing displaying, on a digital display, the one or more weather forecasts.

### Claim 22

A computer-implemented method for forecasting weather conditions and improving agricultural operation, comprising:
obtaining, by a processor, an ensemble of weather models for predicting weather conditions at a field;
obtaining, by the processor, output data from executing the ensemble of weather models for a past period of time;
receiving observed weather data for the past period of time;
generating, by the processor, calibrating data from the output data and the observed weather data using machine learning techniques;
updating the ensemble of weather models to an ensemble of updated weather models based on the calibrating data;
executing the ensemble of updated weather models for a future time to generate a list of weather forecasts for the field;
computing a probability distribution function (PDF) based on the list of weather forecasts;
causing displaying a stacked bar graph based on the PDF.

### Claim 23 (depends on 22)

The computer-implemented method of claim 22, the generating comprising matching up the output data with verifying observations over archived cases.

### Claim 24 (depends on 22)

The computer-implemented method of claim 22, at least one of the ensemble weather models generating a prediction related to precipitation, wind, or temperature.

### Claim 25 (depends on 24)

The computer-implemented method of claim 24, the generating comprises comparing precipitation estimates with a rain gauge or sensor value.

### Claim 26 (depends on 22)

The computer-implemented method of claim 22, the executing the ensemble of updated weather models comprising receiving real-time data as input data to the updated weather models.

### Claim 27 (depends on 22)

The computer-implemented method of claim 22, further comprising causing a display of field-specific calibrated forecasts of the list of weather forecasts with field-specific recently observed weather data.

### Claim 28 (depends on 27)

The computer-implemented method of claim 27, further comprising updating the display at a predetermined frequency.

### Claim 29 (depends on 22)

The computer-implemented method of claim 22, further comprising blending field-specific calibrated forecasts of the list of weather forecasts with field-specific recently observed data to cover gaps in the output data.

### Claim 30 (depends on 22)

The computer-implemented method of claim 22, the updating comprising identifying differences between the observed weather data and the output data at a specific field location over a long history of matchings.

### Claim 31 (depends on 22)

The computer-implemented method of claim 22, the observed weather data being received from in-field or remote sensors, satellites, or lasers.

### Claim 32 (depends on 22)

The computer-implemented method of claim 22, further comprising:
detecting a condition for a trend in the output data;
adjusting the output data according to the trend when the condition is satisfied.

### Claim 33 (depends on 22)

The computer-implemented method of claim 22, the observed weather data including observed precipitation data being rain-gauge based quantitative precipitation estimation (QPE).

### Claim 34 (depends on 22)

The computer-implemented method of claim 22, the machine learning techniques including a quantile regression forest.

### Claim 35 (depends on 22)

The computer-implemented method of claim 22, the stacked bar graph comprising bars being stacked in one direction and corresponding to increasing percentiles, with an axis along that direction corresponding to an increasing amount of precipitation.

### Claim 36 (depends on 22)

The computer-implemented method of claim 22, the stacked bar graph comprising bars being stacked in one direction and corresponding to different ranges of odds, including two identical ranges of odds, with an axis along that direction corresponding an increasing amount of precipitation.

### Claim 37 (depends on 22)

The computer-implemented method of claim 22, the stacked bar graph comprising bars being stacked in one direction and corresponding to decreasing odds of reaching a corresponding amount of precipitation, with an axis along that direction corresponding to an increasing amount of precipitation.

### Claim 38

One or more non-transitory computer-readable media storing instructions which, when executed by one or more processors, cause the one or more processors to perform a method, the method comprising:
obtaining an ensemble of weather models for predicting weather conditions at a field;
obtaining output data from executing the ensemble of weather models for a past period of time;
receiving observed weather data for the past period of time;
generating, by the processor, calibrating data from the output data and the observed weather data using machine learning techniques;
updating the ensemble of weather models to an ensemble of updated weather models based on the calibrating data;
executing the ensemble of updated weather models for a future time to generate a list of weather forecasts for the field;
computing a probability distribution function (PDF) based on the list of weather forecasts;
causing displaying a stacked bar graph based on the PDF.

### Claim 39 (depends on 38)

The one or more non-transitory computer-readable media of claim 38, the updating comprising identifying differences between the observed weather data and the output data at a specific field location over a long history of matchings.

### Claim 40 (depends on 38)

The one or more non-transitory computer-readable media of claim 38, the method further comprising:
detecting a condition for a trend in the output data;
adjusting the output data according to the trend when the condition is satisfied.
