Cultivated land water holding capacity assessment method based on soil moisture memorability

Abstract

The invention relates to a cultivated land water holding capacity assessment method, in particular to a cultivated land water holding capacity assessment method based on soil moisture memorability. The invention aims to solve the problem that the existing method for evaluating the water holding capacity of the cultivated land is difficult to expand to long-time-sequence large-range monitoring. The method comprises the following steps: 1, collecting daily soil moisture time sequence data and terrain classification data of a single pixel of cultivated land; 2, identifying a cultivated land area according to the terrain classification data, and visually interpreting and judging whether the cultivated land is a paddy field or a dry field according to remote sensing images in recent five years; 3, extracting daily soil moisture time sequence data of a single pixel of the SMAP, and determining descending trend data according to a descending trend determination rule; 4, exponential function fitting is carried out on the trend of more than four data pairs in the descending trend; and 5, the water holding capacity of the cultivated land is calculated. According to the invention, a new view angle and method are provided for the evaluation of the water holding capacity of the cultivated land, and technical support is provided for the long-time-sequence large-range monitoring of the water holding capacity of the cultivated land. The invention belongs to the field of cultivated land water holding capacity evaluation.

Description

Technical Field

[0001] The invention relates to a method for evaluating the water holding capacity of cultivated land. Background Art

[0002] Current methods for assessing the water-holding capacity of cultivated land primarily include the specific water capacity index method, the water characteristic curve method, and the cumulative water loss rate method. These methods typically rely on field sampling and laboratory measurements, requiring significant human and material resources and making them difficult to scale up for long-term, large-scale monitoring. Soil water memory methods are often used to study the impact of large-scale, long-term soil moisture variations on climate. However, factors that influence soil water memory, such as soil texture, soil porosity, and soil organic matter, are also linked to the water-holding capacity of cultivated land. Summary of the Invention

[0003] The purpose of the present invention is to solve the technical problem that the existing methods for evaluating the water holding capacity of cultivated land are difficult to expand to large-scale monitoring over a long time series, and to provide a method for evaluating the water holding capacity of cultivated land based on soil moisture memory.

[0004] The method for evaluating the water holding capacity of cultivated land based on soil moisture memory is as follows:

[0005] 1. Based on the long-term daily soil moisture satellite products and land feature classification data, collect daily soil moisture time series data and land feature classification data for single pixels of cultivated land;

[0006] Second, identify cultivated land areas based on land feature classification data, and then determine whether they are paddy fields or dry fields based on visual interpretation of remote sensing images from the past five years;

[0007] Third, extract the daily soil moisture time series data of a single SMAP pixel and determine the downward trend data according to the downward trend determination rule;

[0008] Downtrend determination rules:

[0009] (1) First, sort the daily soil moisture time series data in ascending order according to the DOY (day of year) data, and then calculate the adjacent soil moisture differences in ascending order to obtain the SMD series data;

[0010] Soil moisture difference calculation method:

[0011] Assuming that the current day is day B, the soil moisture value on day B minus the soil moisture value on day B-1 is the soil moisture difference, which is recorded as SMD;

[0012] Then the SMD sequence data is judged to be less than zero. When the SMD is less than zero, that is, the soil moisture value on day B minus the soil moisture value on day B-1 is less than zero, it means that the soil moisture is showing a downward trend at this time, and the downward trend judgment procedure is entered;

[0013] (2) When SMD is greater than 0, assuming that the day is M, it means that the soil moisture value on the Mth day is greater than the soil moisture value on the M-1th day. However, the tolerable error of soil moisture is between 1% and 3%. Therefore, it is determined whether the soil moisture value on the M+1th day is greater than the soil moisture value on the M-1th day. Therefore, TD is calculated, that is, the soil moisture value on the M+1th day minus the soil moisture value on the M-1th day, and it is determined whether TD is less than 0.02;

[0014] If TD is less than 0.02, the soil moisture value on the Mth day is considered to be an abnormal value, so the SMD is determined to be less than 0 from the M+1th day.

[0015] If TD is greater than 0.02, the downward trend is judged to be over, and the duration of the soil moisture downward trend from the Bth day to the Mth day is the number of days, and the soil moisture value and corresponding DOY data of this downward trend are extracted;

[0016] (3) After a period of downward trend determination is completed, continue to search for the time point where the SMD is less than zero, and repeat (2) until all daily soil moisture time series data are determined;

[0017] 4. For the downward trend with more than 4 data pairs, an exponential function is fitted. One data pair represents a soil moisture and corresponding DOY data. The exponential function fitting process is as follows:

[0018] y=a+b*e c*x Formula 1

[0019] Where y is soil moisture; a, b, and c are fitting coefficients; x is DOY data; e is a natural constant, which is the base of the natural logarithm function and an infinite non-repeating decimal, approximately 2.718281828459045;

[0020] 5. Once the exponential function fitting is completed, the soil moisture change model is constructed. The 1 / e of the initial soil moisture of each downward trend is substituted into the soil moisture change model, and the time required for the soil moisture in the downward trend to drop to 1 / e of the initial soil moisture, that is, the soil moisture memory time, is calculated. The average of multiple results is taken, and this average value represents the water holding capacity of the cultivated land. When the soil moisture drops to 1 / e of the initial soil moisture, it means that the soil moisture state at this time is unrelated to the initial state, which means that the soil at this time can no longer return to its original state. The longer the soil moisture memory time, the stronger the water holding capacity of the cultivated land, and vice versa.

[0021] The soil moisture satellite product described in step 1 is the soil moisture product with a spatial resolution of 9 kilometers from the Soil Moisture Active and Passive (SMAP) program.

[0022] The land feature classification data mentioned in step 1 are ESRI Land Use and Land Cover (ESRI LULC 2020) and European Space Agency Land Use and Land Cover (ESA v100).

[0023] The cultivated land in step 1 is a paddy field or a dry field.

[0024] The remote sensing image mentioned in step 2 is Sentinel-2.

[0025] The present invention constructs an evaluation index for cultivated land water holding capacity based on soil moisture memory, and can use remote sensing technology to quickly evaluate cultivated land water holding capacity.

[0026] The present invention is a new method for evaluating the water-holding capacity of cultivated land. It adopts a new perspective of soil moisture memory analysis. According to the prior knowledge of the soil moisture memory method, when the soil moisture reaches 1 / e of the initial soil moisture, it means that the soil at this time can no longer return to its original state. The time required for the original soil moisture to drop to 1 / e of the original soil moisture is called the soil moisture memory time. Therefore, the present invention extends the soil moisture memory time, uses the soil moisture decline trend judgment rule to screen out the moisture retention state of cultivated land in its natural state, and uses the soil moisture memory time to characterize the water-holding capacity of cultivated land. The present invention characterizes the water-holding capacity of cultivated land by the time it takes for the soil moisture to drop to 1 / e of the original soil moisture. The longer the time, the stronger the water-holding capacity of the cultivated land, and vice versa.

[0027] This invention brings a new perspective and method to the assessment of cultivated land water holding capacity, and provides technical support for long-term and large-scale monitoring of cultivated land water holding capacity. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 Schematic diagram of the process of the method for evaluating the water holding capacity of cultivated land based on soil moisture memory of the present invention;

[0029] Figure 2 This is the daily soil moisture time series data graph of a single SMAP pixel extracted in step 3 of experiment 1;

[0030] Figure 3 This is the result chart of the downward trend determination in step 3 of Experiment 1;

[0031] Figure 4 This is the exponential function fitting result graph of step 4 in experiment 1;

[0032] Figure 5 is the water holding capacity of the cultivated land in step five of Experiment 1. The X symbol in the figure represents the number of days corresponding to the drop to 1 / e of the initial soil moisture. DETAILED DESCRIPTION

[0033] The technical solution of the present invention is not limited to the specific embodiments listed below, but also includes any combination of the specific embodiments.

[0034] Specific implementation method 1: This implementation method is based on the soil moisture memory to evaluate the water holding capacity of cultivated land as follows:

[0035] 1. Based on the long-term daily soil moisture satellite products and land feature classification data, collect daily soil moisture time series data and land feature classification data for single pixels of cultivated land;

[0036] Second, identify cultivated land areas based on land feature classification data, and then determine whether they are paddy fields or dry fields based on visual interpretation of remote sensing images from the past five years;

[0037] Third, extract the daily soil moisture time series data of a single SMAP pixel and determine the downward trend data according to the downward trend determination rule;

[0038] Downtrend determination rules:

[0039] (1) First, sort the daily soil moisture time series data in ascending order according to the DOY (day of year) data, and then calculate the adjacent soil moisture differences in ascending order to obtain the SMD series data;

[0040] Soil moisture difference calculation method:

[0041] Assuming that the current day is day B, the soil moisture value on day B minus the soil moisture value on day B-1 is the soil moisture difference, which is recorded as SMD;

[0042] Then the SMD sequence data is judged to be less than zero. When the SMD is less than zero, that is, the soil moisture value on day B minus the soil moisture value on day B-1 is less than zero, it means that the soil moisture is showing a downward trend at this time, and the downward trend judgment procedure is entered;

[0043] (2) When SMD is greater than 0, assuming that the day is M, it means that the soil moisture value on the Mth day is greater than the soil moisture value on the M-1th day. However, the tolerable error of soil moisture is between 1% and 3%. Therefore, it is determined whether the soil moisture value on the M+1th day is greater than the soil moisture value on the M-1th day. Therefore, TD is calculated, that is, the soil moisture value on the M+1th day minus the soil moisture value on the M-1th day, and it is determined whether TD is less than 0.02;

[0044] If TD is less than 0.02, the soil moisture value on the Mth day is considered to be an abnormal value, so the SMD is determined to be less than 0 from the M+1th day.

[0045] If TD is greater than 0.02, the downward trend is judged to be over, and the duration of the soil moisture downward trend from the Bth day to the Mth day is the number of days, and the soil moisture value and corresponding DOY data of this downward trend are extracted;

[0046] (3) After a period of downward trend determination is completed, continue to search for the time point where the SMD is less than zero, and repeat (2) until all daily soil moisture time series data are determined;

[0047] 4. For the downward trend with more than 4 data pairs, an exponential function is fitted. One data pair represents a soil moisture and corresponding DOY data. The exponential function fitting process is as follows:

[0048] y=a+b*e c*x

[0049] Formula 1

[0050] Where y is soil moisture; a, b, and c are fitting coefficients; x is DOY data; e is a natural constant, which is the base of the natural logarithm function and an infinite non-repeating decimal, approximately 2.718281828459045;

[0051] 5. Once the exponential function fitting is completed, the soil moisture change model is constructed. The 1 / e of the initial soil moisture of each downward trend is substituted into the soil moisture change model, and the time required for the soil moisture in the downward trend to drop to 1 / e of the initial soil moisture, that is, the soil moisture memory time, is calculated. The average of multiple results is taken, and this average value represents the water holding capacity of the cultivated land. When the soil moisture drops to 1 / e of the initial soil moisture, it means that the soil moisture state at this time is unrelated to the initial state, which means that the soil at this time can no longer return to its original state. The longer the soil moisture memory time, the stronger the water holding capacity of the cultivated land, and vice versa.

[0052] Specific embodiment 2: This embodiment differs from specific embodiment 1 in that the soil moisture satellite product in step 1 is a soil moisture product with a spatial resolution of 9 kilometers from the Soil Moisture Active and Passive (SMAP) program. Other aspects are the same as specific embodiment 1.

[0053] Specific embodiment 3: This embodiment differs from specific embodiments 1 or 2 in that the feature classification data in step 1 are ESRI Land Use and Land Cover (ESRI LULC 2020) and European Space Agency Land Use and Land Cover (ESA v100). Other aspects are the same as specific embodiments 1 or 2.

[0054] Specific embodiment 4: This embodiment differs from specific embodiments 1 to 3 in that the cultivated land in step 1 is a paddy field or dry field. Other aspects are the same as specific embodiments 1 to 3.

[0055] Specific embodiment 5: This embodiment differs from specific embodiments 1 to 4 in that the remote sensing image in step 2 is Sentinel-2. Otherwise, it is the same as specific embodiments 1 to 4. The following experiment was used to verify the effect of the present invention:

[0056] Experiment 1:

[0057] The method for evaluating the water holding capacity of cultivated land based on soil moisture memory is as follows:

[0058] 1. Based on the long-term daily soil moisture satellite products and land feature classification data, collect the daily soil moisture time series data and land feature classification data of single pixel of cultivated land (Appendix Figure 2-5 The data is sourced from the China Network on the ISMN website (https: / / ismn.earth / en / ), extracted based on the latitude and longitude of the Youyi site. The year used in this article is 2017.

[0059] Second, identify cultivated land areas based on land feature classification data, and then determine whether they are paddy fields or dry fields based on visual interpretation of remote sensing images from the past five years;

[0060] Third, extract the daily soil moisture time series data of a single SMAP pixel and determine the downward trend data according to the downward trend determination rule;

[0061] Downtrend determination rules:

[0062] (1) First, sort the daily soil moisture time series data in ascending order according to the DOY (day of year) data, and then calculate the adjacent soil moisture differences in ascending order to obtain the SMD series data;

[0063] Soil moisture difference calculation method:

[0064] Assuming that the current day is day B, the soil moisture value on day B minus the soil moisture value on day B-1 is the soil moisture difference, which is recorded as SMD;

[0065] Then the SMD sequence data is judged to be less than zero. When the SMD is less than zero, that is, the soil moisture value on day B minus the soil moisture value on day B-1 is less than zero, it means that the soil moisture is showing a downward trend at this time, and the downward trend judgment procedure is entered;

[0066] (2) When SMD is greater than 0, assuming that the day is M, it means that the soil moisture value on the Mth day is greater than the soil moisture value on the M-1th day. However, the tolerable error of soil moisture is between 1% and 3%. Therefore, it is determined whether the soil moisture value on the M+1th day is greater than the soil moisture value on the M-1th day. Therefore, TD is calculated, that is, the soil moisture value on the M+1th day minus the soil moisture value on the M-1th day, and it is determined whether TD is less than 0.02;

[0067] If TD is less than 0.02, the soil moisture value on the Mth day is considered to be an abnormal value, so the SMD is determined to be less than 0 from the M+1th day.

[0068] If TD is greater than 0.02, the downward trend is judged to be over, and the duration of the soil moisture downward trend from the Bth day to the Mth day is the number of days, and the soil moisture value and corresponding DOY data of this downward trend are extracted;

[0069] (3) After a period of downward trend determination is completed, continue to search for the time point where the SMD is less than zero, and repeat (2) until all daily soil moisture time series data are determined;

[0070] 4. For the downward trend with more than 4 data pairs, an exponential function is fitted. One data pair represents a soil moisture and corresponding DOY data. The exponential function fitting process is as follows:

[0071] y=a+b*e c*x Formula 1

[0072] Where y is soil moisture; a, b, and c are fitting coefficients; x is DOY data; e is a natural constant, which is the base of the natural logarithm function and an infinite non-repeating decimal, approximately 2.718281828459045;

[0073] 5. After the exponential function fitting is completed, a soil moisture variation model is constructed. 1 / e of the initial soil moisture value for each downward trend is substituted into the soil moisture variation model. The time required for the soil moisture to drop to 1 / e of the initial soil moisture value during the downward trend is calculated, i.e., the soil moisture memory duration. Multiple results are averaged, and this average value represents the water-holding capacity of the cultivated land. The final output is an average water-holding capacity value of 23. When the soil moisture drops to 1 / e of the initial soil moisture value, the soil moisture state at this point is unrelated to the initial state, indicating that the soil at this point can no longer return to its initial state. A longer soil moisture memory duration indicates a stronger water-holding capacity, and vice versa.

[0074] The soil moisture satellite product described in step 1 is the soil moisture product with a spatial resolution of 9 kilometers from the Soil Moisture Active and Passive (SMAP) program.

[0075] The land feature classification data mentioned in step 1 are ESRI Land Use and Land Cover (ESRI LULC 2020) and European Space Agency Land Use and Land Cover (ESA v100).

[0076] The cultivated land in step 1 is a paddy field or a dry field.

[0077] The remote sensing image mentioned in step 2 is Sentinel-2.

Claims

1. A method for evaluating the water holding capacity of cultivated land based on soil moisture memory, characterized by The method for evaluating the water holding capacity of cultivated land based on soil moisture memory is as follows:

1. Based on the long-term daily soil moisture satellite products and land feature classification data, collect daily soil moisture time series data and land feature classification data for single pixels of cultivated land; Second, identify cultivated land areas based on land feature classification data, and then determine whether they are paddy fields or dry fields based on visual interpretation of remote sensing images from the past five years; 3. Extract daily soil moisture time series data of a single pixel and determine the downward trend data according to the downward trend determination rule; Downtrend determination rules: (1) First, the daily soil moisture time series data are sorted in ascending order according to the DOY data, and then the adjacent soil moisture differences are calculated in ascending order to obtain the SMD series data; Soil moisture difference calculation method: Assuming that the current day is day B, the soil moisture value on day B minus the soil moisture value on day B-1 is the soil moisture difference, which is recorded as SMD; Then the SMD sequence data is judged to be less than zero. When the SMD is less than zero, that is, the soil moisture value on day B minus the soil moisture value on day B-1 is less than zero, it means that the soil moisture is showing a downward trend at this time, and the downward trend judgment procedure is entered; (2) When SMD is greater than 0, assuming that the day is M, it means that the soil moisture value on the Mth day is greater than the soil moisture value on the M-1th day. However, the tolerable error of soil moisture is between 1% and 3%. Therefore, it is determined whether the soil moisture value on the M+1th day is greater than the soil moisture value on the M-1th day. Therefore, TD is calculated, that is, the soil moisture value on the M+1th day minus the soil moisture value on the M-1th day, and it is determined whether TD is less than 0.02; If TD is less than 0.02, the soil moisture value on the Mth day is considered to be an abnormal value, so the SMD is determined to be less than 0 from the M+1th day. If TD is greater than 0.02, the downward trend is judged to be over, and the duration of the soil moisture downward trend from the Bth day to the Mth day is the number of days, and the soil moisture value and corresponding DOY data of this downward trend are extracted; (3) After a period of downward trend determination is completed, continue to search for the time point where the SMD is less than zero, and repeat (2) until all daily soil moisture time series data are determined; 4. For the downward trend with more than 4 data pairs, an exponential function is fitted. One data pair represents a soil moisture and corresponding DOY data. The exponential function fitting process is as follows: y=a+b*e c*x Formula 1 Where y is soil moisture; a, b, c are fitting coefficients; x is DOY data; e is a natural constant; 5. Once the exponential function fitting is completed, the soil moisture change model is constructed. The 1 / e of the initial soil moisture of each downward trend is substituted into the soil moisture change model, and the time required for the soil moisture in the downward trend to drop to 1 / e of the initial soil moisture, that is, the soil moisture memory time, is calculated. The average of multiple results is taken, and this average value represents the water holding capacity of the cultivated land. When the soil moisture drops to 1 / e of the initial soil moisture, it means that the soil moisture state at this time is unrelated to the initial state, which means that the soil at this time can no longer return to its original state. The longer the soil moisture memory time, the stronger the water holding capacity of the cultivated land, and vice versa.

2. The method for evaluating the water holding capacity of cultivated land based on soil moisture memory according to claim 1, characterized in that The soil moisture satellite product described in step 1 is a soil moisture product with a spatial resolution of 9 kilometers in the soil moisture active and passive detection program.

3. The method for evaluating the water holding capacity of cultivated land based on soil moisture memory according to claim 1, characterized in that The land feature classification data mentioned in step 1 are the land use and land cover data of Yizhirui and the land use and land cover data of the European Space Agency.

4. The method for evaluating the water holding capacity of cultivated land based on soil moisture memory according to claim 1, characterized in that The cultivated land in step 1 is a paddy field or a dry field.

5. The method for evaluating the water holding capacity of cultivated land based on soil moisture memory according to claim 1, characterized in that The remote sensing image mentioned in step 2 is Sentinel-2.

Images