Method and device for generating a spatiotemporal map of an agriculturally relevant physical quantity, as well as method and device for controlling an agricultural work process

Abstract

The invention relates to a method for generating a spatiotemporal map (119) of an agriculturally relevant physical quantity. The method comprises a step of reading first measurement data (M1) from a mobile sensor (M) and second measurement data (S2) from a stationary sensor (S) via an input interface (101). The mobile sensor (M) is arranged on an agricultural machine (D) that moves within a predefined agricultural work area (A). The stationary sensor (S) is arranged in a fixed geographical relationship to the work area (A). The method further comprises a step of determining correlation data (115) using the first measurement data (M1) and the second measurement data (S2), wherein the correlation data (115) represent a correlation between the first measurement data (M1) and the second measurement data (S2).The procedure also includes a step of generating virtual measurements using the correlation data (115) and the second measurement data (S2) to generate the spatiotemporal map (119) using the first measurements (M1), the second measurements (S2) and the virtual measurements.

Description

State of the art

[0001] The present invention relates to a method for generating a spatial-temporal map of an agriculturally relevant physical quantity, to a method for controlling an agricultural work process, to a corresponding device and to a corresponding computer program product and a corresponding machine-readable storage medium.

[0002] Environmental conditions, such as the weather, can have a strong influence on agricultural processes and determine many of the decisions that need to be made, particularly during crop cultivation. Traditionally, current weather data and forecasts can be compiled and made available based on stationary weather stations, which may be operated by government agencies or private companies, as well as satellite data. The distance between weather stations can typically be relatively large, for example, several kilometers or tens of kilometers. More local information can be obtained, in particular, by operating a weather station on one's own property and storing the data. Some of this data may also be publicly available.

[0003] US 2016 / 0116640 A describes a method for predicting local weather conditions based on current satellite weather data and a forecast model. The forecast model represents a correlation between historical satellite weather data and historical local weather measurements generated using smartphone sensors. Disclosure of the invention

[0004] Against this background, the approach presented here presents a method for generating a spatial-temporal map of an agriculturally relevant physical quantity, a method for controlling an agricultural work process, a device that uses one of these methods, and finally, a corresponding computer program product and a corresponding machine-readable storage medium according to the main claims. Advantageous embodiments emerge from the respective subclaims and the following description.

[0005] According to embodiments, in particular, a virtual sensor or a prediction method for sparse spatial-temporal data can be provided for agricultural applications. For example, a prediction method or a virtual sensor can be provided that can deliver dense data based on sparse spatial-temporal data measured by a real sensor. This can be used, for example, to create microclimate maps that can serve as a basis for decisions regarding agricultural processes. Unlike in the prior art, for example, satellite weather data can be dispensed with, whereby the same measured variables can be directly recorded in a stationary manner as using mobile sensors, thus eliminating the need for an additional model for comparing the measured data. Derivation of historical data, which is essential for agricultural applications, can be realized.

[0006] A method for generating a spatial-temporal map of an agriculturally relevant physical quantity is presented, comprising the following steps: Reading in first measurement data from a mobile sensor and second measurement data from a stationary sensor via an input interface, wherein the mobile sensor is arranged on an agricultural work machine that moves in a predefined agricultural work area, wherein the stationary sensor is arranged in a fixed geographical relationship to the work area, wherein the first measurement data represent a plurality of first measured values ​​of the physical quantity, wherein each first measured value is assigned a measurement time and a geographical measurement position within the work area, wherein the second measurement data represent a plurality of second measured values ​​of the physical quantity measured at the times of the first measured values; Determining correlation data using the first measurement data and the second measurement data, wherein the correlation data represents a correlation between the first measurement data and the second measurement data; and Generating virtual measurements using the correlation data and the second measurement data to generate the spatio-temporal map using the first measurement values, the second measurement values ​​and the virtual measurement values.

[0007] The method may be a computer-implemented method. In the generation step, the virtual measured values ​​may be generated using a determination rule, which may, for example, include a prediction of time series such as autoregressive integrated moving averages (ARIMA), matrix factorization models, and other statistical models such as genetic models or vector machines.

[0008] This allows a spatial or spatio-temporal map of a physical quantity, e.g., temperature, to be determined based on measurement data from a mobile sensor, e.g., a temperature sensor on a work machine, and measurement data from a stationary sensor, e.g., a fixed temperature sensor at the edge of a field. For this purpose, a correlation can be determined between the measurement data of the mobile and stationary sensors. The determined correlation and the measurement data of the stationary sensor can then be used, in particular, to calculate the values ​​of the physical quantity at the measurement positions of the mobile sensor for times when the mobile sensor is not at the respective measurement position.

[0009] This makes it possible, for example, to place individual measurements at a specific location and time into a context that allows for spatiotemporally extended operations. In this way, a detailed spatiotemporal map can be created from sparse measurements in space, enabling more precise agricultural planning and action. In particular, precise knowledge of past, current, and future environmental conditions can be used for agricultural processes to adapt crop management and achieve expected yields.

[0010] According to one embodiment, in the generation step, each virtual measured value can be assigned a virtual measurement time that is different from the measurement times and one of the geographical measurement positions. Thus, values ​​of the physical quantity at the measurement positions of the mobile sensor can be advantageously determined for times at which the mobile sensor is not, is not yet, or is no longer located at the respective measurement position. Thus, the representation of the measured values ​​in the map can be condensed over time.

[0011] During the generation step, each virtual measured value can also be assigned one of the measurement times and a virtual geographical measurement position within the working area that is different from the measurement positions. This allows the representation of the measured values ​​in the map to be spatially densified or, in other words, a finer spatial resolution of the measured values ​​to be generated.

[0012] Likewise, during the generation step, each virtual measured value can be assigned a virtual measurement time point that is different from the measurement times and a virtual geographical measurement position within the working area that is different from the measurement positions. This embodiment offers the advantage that the representation of the measured values ​​on the map can be refined and supplemented as needed.

[0013] Furthermore, in the determining step, a difference can be determined between each first measured value and the second measured value measured at the same time. The plurality of differences thus obtained can be used as the correlation data. A difference can also be referred to as an offset. Thus, a simple and meaningful relationship can be established between the first and second measured values, which can be reliably used to generate the virtual measured values.

[0014] The initial measurements or the differences can be combined. This allows particularly robust correlation data or particularly accurate modeling with respect to other observables to be obtained.

[0015] For a geographical measurement position, the difference can also be a function of time and at least one other physical quantity, and the correlation data can be determined from this function using at least the second measured value as an approximate value. This offers the advantage of simple and reliable determination of the correlation data.

[0016] In addition, in the determination step, the correlation data can be further parameterized using at least one additional physical quantity and, additionally or alternatively, at least one physical model. Such an embodiment offers the advantage that physical relationships between different measured quantities can be taken into account.

[0017] In particular, the measurement data acquired during the reading step can represent measured values ​​of temperature, humidity, wind speed, wind direction, air pressure, soil moisture, solar radiation, particle concentration, precipitation amount, and / or nutrient content as physical quantities. Thus, physical quantities relevant to agriculture, for which only sparse measured values ​​are available, can be advantageously mapped.

[0018] According to one embodiment, the second measurement data read in during the reading step can be read in by a stationary sensor implemented as a physical sensor. Alternatively, the second measurement data read in during the reading step can be read in by a stationary sensor implemented as a virtual sensor. The virtual sensor can be modeled using the first measurement data by at least one statistical rule, in particular averaging over an entire measurement period or a moving box average, and additionally or alternatively by at least one functional rule, in particular linear interpolation or higher-order functions based on physical processes over time. Thus, a suitable stationary sensor can be used depending on the specific conditions in the field of work.

[0019] A method for controlling an agricultural work process is also presented, the method comprising the following steps: Generating a spatial-temporal map of an agriculturally relevant physical quantity according to an embodiment of a method for generating mentioned herein; Evaluating the map to determine the extent to which the physical quantity in the work area fulfils a process condition for the agricultural work process; and Outputting a control signal to an output interface to at least one agricultural work machine or planning device for carrying out the agricultural work process depending on the evaluation result.

[0020] The method can be a computer-implemented method. The agricultural work process can be, for example, a process for soil cultivation, sowing, plant treatment, or harvesting. When executing the generating step, the steps of an embodiment of a generating method mentioned herein can be executed as substeps. The evaluation result can, for example, indicate whether a temperature sum has exceeded a threshold value that represents a flowering time or harvest time of an agricultural crop.

[0021] According to one embodiment, the generating step can be performed repeatedly for the same physical quantity to generate an updated map. Additionally or alternatively, the generating step can be performed repeatedly for different physical quantities to generate multiple maps for different physical quantities. In this case, the updated map and, additionally or alternatively, the multiple maps can be evaluated in the evaluating step. Such an embodiment offers the advantage that a particularly meaningful data basis with a particularly precise and robust correlation between the first and second measured values ​​can be used.

[0022] The approach presented here further provides a device designed to carry out or implement the steps of a variant of a method presented here in corresponding devices. This embodiment of the invention in the form of a device also allows the problem underlying the invention to be solved quickly and efficiently.

[0023] In this case, a device can be understood as an electrical device that processes sensor signals and outputs control and / or data signals depending on them. The device can have an interface, which can be implemented in hardware and / or software. In a hardware implementation, the interfaces can, for example, be part of a so-called system ASIC, which contains a wide variety of functions of the device. However, it is also possible for the interfaces to be separate integrated circuits or to consist at least partially of discrete components. In a software implementation, the interfaces can be software modules that are present, for example, on a microcontroller or control unit alongside other software modules.

[0024] Also advantageous is a computer program product with program code that can be stored on a machine-readable medium or a machine-readable carrier such as a semiconductor memory, a hard disk memory or an optical memory and is used to carry out the methods according to one of the embodiments described above when the program product is executed on a computer or a device.

[0025] The approach presented here is explained in more detail below using the attached drawings. They show: Fig. 1 is a schematic representation of a device according to an embodiment of the present invention; Fig. 2 shows a flow diagram of an embodiment of a method for generating a spatial-temporal map of an agriculturally relevant physical quantity; and Fig. 3 a flowchart of an embodiment of a method for controlling an agricultural work process.

[0026] In the following description of advantageous embodiments of the present invention, the same or similar reference numerals are used for the elements shown in the various figures and having a similar effect, whereby a repeated description of these elements is omitted.

[0027] Fig. 1 shows a schematic representation of a device 100 according to an embodiment of the present invention. The device 100 is designed to generate a spatial-temporal map of an agriculturally relevant physical quantity and / or to control an agricultural work process. For this purpose, the device 100 comprises a generation unit 110 and / or a control unit 120. The device 100 also comprises an input interface 101 and an output interface 103. The generation unit 110 and the control unit 120 of the device 100 will be discussed in more detail below.

[0028] In Fig. 1 shows a predefined agricultural work area A, for example a field, arable land or the like. An agricultural work machine D moves in the work area A. Furthermore, another agricultural work machine E is shown. A mobile sensor M is arranged on the work machine D. The mobile sensor M is designed to measure at least one agriculturally relevant physical quantity and to provide it as first measured values ​​in the form of first measurement data M1. A stationary sensor S is arranged in a fixed geographical relationship to the work area A. The stationary sensor S is designed to measure at least one agriculturally relevant physical quantity and to provide it as second measured values ​​in the form of second measurement data S2.

[0029] An agriculturally relevant physical quantity includes, for example, temperature, air humidity, wind speed, wind direction, air pressure, soil moisture, solar radiation, particle concentration, precipitation amount and / or nutrient content.

[0030] The generation unit 110 of the device 100 comprises a reading device 112, a determination device 114, and a generation device 116. The reading device 112 is designed to read in the first measurement data M1 from the mobile sensor M via the input interface 101 of the device 100. The first measurement data M1 represents a plurality of the first measured values ​​of the physical quantity. Each first measured value M1 is assigned a measurement time and a geographical measurement position B within the work area A. The reading device 112 is also designed to read in the second measurement data S2 from the stationary sensor S via the input interface 101 of the device 100. The second measurement data S2 represents a plurality of the second measured values ​​of the physical quantity measured at the times of the first measured values.Furthermore, the reading device 112 is designed to forward the first measurement data M1 and the second measurement data S2 to the determination device 114.

[0031] According to one embodiment, the input device 112 is configured to input the second measurement data S2 from a stationary sensor S, which is implemented as a physical sensor or as a virtual sensor. In the case of a virtual sensor, the sensor is modeled using the first measurement data M1 by at least one statistical rule, in particular averaging over an entire measurement period or a moving box average, and / or by at least one functional rule, in particular linear interpolation or higher-order functions based on physical processes over time.

[0032] The determination device 114 of the device 100, more precisely the generation unit 110 of the device 100, is configured to determine correlation data 115 using the first measurement data M1 and the second measurement data S2. The correlation data 115 represents a correlation between the first measurement data M1 and the second measurement data S2. The determination device 114 is also configured to forward the correlation data 115, the first measurement data M1, and the second measurement data S2 to the generation device 116.

[0033] According to one exemplary embodiment, the determination device 114 is designed to determine a difference between each first measured value M1 and the second measured value S2 measured at the same time. The determination device 114 is designed to use the plurality of differences or offsets thus obtained as the correlation data 115. In particular, the determination device 114 is designed to combine the first measured values ​​M1 or the differences or offsets. Additionally or alternatively, the determination device 114 is designed in particular to determine the difference for a geographical measurement position B as a function of time and of at least one further physical quantity and to determine the correlation data 115 from this function using at least the second measured value S2 as an approximate value.According to one embodiment, the determination device 114 is designed to further parameterize the correlation data 115 using at least one further physical quantity and / or at least one physical model.

[0034] The generation device 116 is configured to generate virtual measurement values ​​using the correlation data 115 and the second measurement data S2 in order to generate the spatio-temporal map in the form of map data 119 using the first measurement values ​​M1, the second measurement values ​​S2, and the virtual measurement values. Thus, the generation device 116 is also configured to generate the map data 119 using the first measurement values ​​M1, the second measurement values ​​S2, and the generated virtual measurement values.

[0035] According to one embodiment, the generating device 116 is designed to generate the virtual measured values ​​such that each virtual measured value is assigned a virtual measurement time different from the measurement times and one of the geographical measurement positions B. Additionally or alternatively, the generating device 116 is designed to generate the virtual measured values ​​such that each virtual measured value is assigned one of the measurement times and a virtual geographical measurement position C different from the measurement positions B within the work area A. Additionally or alternatively, the generating device 116 is designed to generate the virtual measured values ​​such that each virtual measured value is assigned a virtual measurement time different from the measurement times and a virtual geographical measurement position C different from the measurement positions B within the work area A.

[0036] The control unit 120 of the device 100 comprises the generating unit 110, an evaluation device 124, and an output device 126. The evaluation device 124 is configured to receive the map data 119 from the generating unit 110, more precisely, the generating device 116 of the generating unit 110. The evaluation device 124 is also configured to evaluate the spatial-temporal map represented by the map data 119 in order to determine, as an evaluation result 125, the extent to which the physical variable in the work area A fulfills a process condition for the agricultural work process to be performed. Furthermore, the evaluation device 124 is configured to forward the evaluation result 125 to the output device 126.The output device 126 is designed to output a control signal 129 to the output interface 103 to at least one agricultural work machine D and / or E or planning device for carrying out the agricultural work process, depending on the evaluation result 125.

[0037] According to one embodiment, the generation unit 110 is configured to repeatedly generate the map for the same physical quantity or, more precisely, to generate an updated map for the same physical quantity. Additionally or alternatively, the generation unit 110 is configured to generate multiple maps for different physical quantities. The evaluation device 124 is configured to evaluate the updated map and / or the multiple maps.

[0038] Fig. Figure 2 shows a flowchart of an embodiment of a method 210 for generating a spatial-temporal map of an agriculturally relevant physical quantity. The method 210 for generating can be carried out using a device or, more precisely, a generating unit of a device similar to that of Fig. 1. The method 210 for generating comprises a step 212 of reading, a step 214 of determining, and a step 216 of generating.

[0039] In step 212 of reading in, measurement data from a mobile sensor and second measurement data from a stationary sensor are read in via an input interface. The mobile sensor is arranged on an agricultural work machine that moves in a predefined agricultural work area. The stationary sensor is arranged in a fixed geographical relationship to the work area. The first measurement data represent a plurality of first measured values ​​of the physical quantity, with each first measured value being assigned a measurement time and a geographical measurement position within the work area. The second measurement data represent a plurality of second measured values ​​of the physical quantity measured at the times of the first measured values.In the determining step 214, correlation data representing a correlation between the first measurement data and the second measurement data is determined using the first measurement data and the second measurement data. In the generating step 216, virtual measurement values ​​are generated using the correlation data and the second measurement data to create the spatio-temporal map using the first measurement values, the second measurement values, and the virtual measurement values.

[0040] Fig. Figure 3 shows a flowchart of an embodiment of a method 320 for controlling an agricultural work process. The method 320 for controlling can be executed using a device or, more precisely, a control unit of a device similar to that of Fig. 1. The method 320 for controlling comprises a generating step 310, an evaluating step 324, and an outputting step 326.

[0041] The step 310 of generating comprises as sub-steps the steps of the method for generating from Fig. 2. Thus, in step 310 of generating, a spatial-temporal map of an agriculturally relevant physical quantity is generated, as described in Fig. 2. In evaluation step 324, the map is evaluated to determine, as the evaluation result, the extent to which the physical quantity in the work area fulfills a process condition for the agricultural work process. In output step 326, depending on the evaluation result, a control signal is output to an output interface to at least one agricultural work machine or planning device for carrying out the agricultural work process.

[0042] The generating step 310 is performed repeatedly for the same physical quantity to generate an updated map. Additionally or alternatively, the generating step 310 is performed repeatedly for different physical quantities to generate multiple maps for different physical quantities. In this case, the updated map and / or the multiple maps are then evaluated in the evaluating step 324.

[0043] With reference to the figures described above, exemplary embodiments are summarized again below and, in other words, briefly explained.

[0044] During a deployment in a field or work area A, a physical quantity is continuously monitored. This could be, for example, temperature, humidity, wind speed, wind direction, air pressure, soil moisture, solar radiation, particle concentration, or precipitation amount. With each measurement using the mobile sensor M, location information, such as latitude, longitude, and possibly altitude, is also recorded in the initial measurement data M1. The measurements can be repeated, for example, at a fixed time interval, e.g., every second, or based on the location after traveling a certain distance, e.g., every 5 meters.

[0045] In the same time frame as the process in the field or work area A, the same quantity(s), which are correlated with each other, are measured by the stationary sensor S as a reference sensor. A reference sensor can be a stationary observation post such as a weather station. Alternatively, the reference sensor can be constructed from the data generated by deployment in the field or work area A. This can be done using statistical methods such as averaging over the entire time period or using a moving box average. It can also be modeled by a function, e.g., by linear interpolation or higher-order functions based on physical processes over time.

[0046] An example of a physical quantity is temperature over the day or year, for example approximated by trigonometric functions. A difference or offset is calculated between the reference sensor value or the second measured value at a specific time and the measured value or first measured value. This offset calculation can be further parameterized by taking into account other measured values ​​and (preferably) physical modeling; for example, the temperature near the ground can be influenced by solar radiation and wind. From the offsets, a fine-grained map of the field or work area A can be created. By referring to the reference sensor or stationary sensor S, the offset values ​​are not just momentary observations, but can be extrapolated over time to predict the local climate. Temperature sums, for example, are used to predict flowering or harvest times of crops.Using offset maps or the spatio-temporal map, this can be calculated with high spatial resolution. Another application example is the mapping of nutrient levels in field or working area A, particularly the nitrogen content.

[0047] For upcoming operations or work processes, the method 210 and / or 320 can be repeated. The various locally measured values ​​or offsets can be combined to obtain a more robust offset or a more accurate modeling with respect to other observables. The offsets at a location can be a function of time and all other available observables. From this, an offset value can be calculated, using at least the reference sensor value or second measured value as an approximate value.

[0048] A temperature-specific example is referred to again below. A mobile sensor M designed as a temperature sensor on an agricultural machine measures the temperature while it is in use in the work area A or field. The temperature is recorded together with the current location of the work machine D as the geographical position B. A nearby reference temperature sensor or stationary sensor S measures the temperature over the same period. The offset or the difference between the reference temperature as the second measured value and the locally measured temperature value as the first measured value is calculated. A map with the offset temperature values ​​is created using the location information. The temperature sum can be used to predict the flowering time of crops in the field or work area A. Instead of measuring the temperature sum only via the reference temperature sensor orInstead of calculating the time of flowering using a stationary sensor S or a coarse-grained forecast, this can now be done using the data offset at each location, including virtual positions C, with the differences being accumulated. This can be used, for example, to obtain a flowering time with high spatial resolution.

[0049] The following also looks at the example of particulate matter (PM) or particle concentration as a physical quantity. A mobile sensor M designed as a PM sensor on an agricultural machine or work machine D measures the particle concentration while it is working in the field or work area A, whereby the particulate matter caused by the work is assumed to be negligible. The concentration is recorded together with the current location or geographical position B of the machine or work machine D. A nearby reference PM sensor or stationary sensor S measures the concentration over the same period. A spatial model with sinks and sources of PM can be refined using this information, since the work machine D provides different distances, while the reference station or stationary sensor S specifies a fixed distance.An increase or decrease in FA concentration propagates from a source at a certain velocity. Observing the increase or decrease can be used to calculate the propagation or, through triangulation, to precisely determine the position of the source(s). This eliminates ambiguities that can arise from purely stationary measurement positions.

[0050] According to embodiments, for example, maps of observed values ​​or physical quantities are generated that show offsets of the observed values. In a common implementation, offsets aggregated over time are used. This enables the creation of fine-grained maps even without an extensive network of physical sensors.

[0051] The exemplary embodiments described and shown in the figures are selected only as examples. Different exemplary embodiments can be combined with each other in their entirety or with regard to individual features. Furthermore, one exemplary embodiment can be supplemented by features of another exemplary embodiment.

[0052] Furthermore, the process steps presented here can be repeated and carried out in a different order than that described.

[0053] If an embodiment comprises an “and / or” link between a first feature and a second feature, this is to be read as meaning that the embodiment according to one embodiment has both the first feature and the second feature and according to another embodiment has either only the first feature or only the second feature. QUOTES CONTAINED IN THE DESCRIPTION

[0000] This list of documents submitted by the applicant was generated automatically and is included solely for the convenience of the reader. This list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions. Cited patent literature

[0000] US 2016 / 0116640 A

[0003]

Claims

[1] Method (210) for generating a spatial-temporal map (119) of an agriculturally relevant physical quantity, the method (210) comprising the following steps: Reading in (212) first measurement data (M1) from a mobile sensor (M) and second measurement data (S2) from a stationary sensor (S) via an input interface (101), wherein the mobile sensor (M) is arranged on an agricultural work machine (D) that moves in a predefined agricultural work area (A), wherein the stationary sensor (S) is arranged in a fixed geographical relationship to the work area (A), wherein the first measurement data (M1) represent a plurality of first measured values ​​of the physical quantity, wherein each first measured value is assigned a measurement time and a geographical measurement position (B) within the work area (A), wherein the second measurement data (S2) represent a plurality of second measured values ​​of the physical quantity measured at the times of the first measured values; Determining (214) correlation data (115) using the first measurement data (M1) and the second measurement data (S2), wherein the correlation data (115) represents a correlation between the first measurement data (M1) and the second measurement data (S2); and Generating (216) virtual measurement values ​​using the correlation data (115) and the second measurement data (S2) to generate the spatial-temporal map (119) using the first measurement values ​​(M1), the second measurement values ​​(S2) and the virtual measurement values. [2] Method (210) according to claim 1, wherein in the step (216) of generating, each virtual measured value is assigned a virtual measurement time different from the measurement times and one of the geographical measurement positions (B). [3] Method (210) according to claim 1, wherein in the step (216) of generating, each virtual measured value is assigned one of the measuring times and a virtual geographical measuring position (C) different from the measuring positions (B) within the working area (A). [4] Method (210) according to claim 1, wherein in the step (216) of generating, each virtual measured value is assigned a virtual measurement time different from the measurement times and a virtual geographical measurement position (C) different from the measurement positions (B) within the working area (A). [5] Method (210) according to one of the preceding claims, wherein in the step (214) of determining a difference is determined between each first measured value and the second measured value measured at the same time, wherein in particular the plurality of differences thus obtained are used as the correlation data (115). [6] Method (210) according to claim 5, wherein the first measured values ​​or the differences are combined. [7] Method (210) according to one of claims 5 to 6, wherein for a geographical measurement position (B) the difference is a function of time and of at least one further physical quantity and the correlation data (115) are determined from this function using at least the second measured value as an approximate value. [8] Method (210) according to one of the preceding claims, wherein in the step (214) of determining, the correlation data (115) are further parameterized using at least one further physical quantity and / or at least one physical modeling. [9] Method (210) according to one of the preceding claims, wherein the measurement data (M1, S2) read in step (212) of reading represent measured values ​​of a temperature, an air humidity, a wind speed, a wind direction, an air pressure, a soil moisture, a solar radiation, a particle concentration, a precipitation amount and / or a nutrient content as a physical quantity. [10] Method (210) according to one of the preceding claims, wherein the second measurement data (S2) read in in step (212) of reading in are read in by a stationary sensor (S) which is implemented as a physical sensor or as a virtual sensor, wherein the virtual sensor is modeled using the first measurement data (M1) by at least one statistical rule, in particular averaging over an entire measurement period or a moving box average, and / or by at least one functional rule, in particular linear interpolation or higher-order functions based on physical processes over time. [11] Method (320) for controlling an agricultural work process, the method (320) comprising the following steps: Generating (310) a spatial-temporal map (119) of an agriculturally relevant physical quantity according to the method (210) for generating according to one of the preceding claims; Evaluating (324) the map (119) to determine as an evaluation result (125) the extent to which the physical quantity in the work area (A) fulfills a process condition for the agricultural work process; and Outputting (326) a control signal (129) to an output interface (103) to at least one agricultural work machine or planning device (E) for carrying out the agricultural work process depending on the evaluation result (125). [12] The method (320) of claim 11, wherein the generating step (310) is repeatedly performed for the same physical quantity to generate an updated map and / or is repeatedly performed for different physical quantities to generate a plurality of maps for different physical quantities, wherein in the evaluating step (324) the updated map and / or the plurality of maps are evaluated. [13] Device (100) which is designed to carry out the steps of a method (210; 320) according to one of claims 1 to 12 in corresponding devices (110; 120; 112, 114, 116; 124, 126). [14] Computer program product with program code for carrying out the method (210; 320) according to one of claims 1 to 12, when the program product is executed on a device (100). [15] A machine-readable storage medium on which the computer program product according to claim 14 is stored.

Images