Drought monitoring method based on comprehensive deviation drought index integrating multiple indicators

Abstract

The present invention discloses a drought monitoring method that integrates a comprehensive deviation drought index (CDDI) with multiple indicators. The method comprises: calculating a monthly average climate dataset for each calendar month during the study phase by coordinating terrestrial water storage (TWS), precipitation (PET), and soil moisture (SM) product data; calculating monthly climate anomalies for terrestrial water storage, precipitation, and soil moisture for month j of year i based on the monthly average climate dataset obtained in step 1; standardizing the monthly climate anomalies calculated in step 2 and constructing a comprehensive deviation (CD) based on the standardized climate anomalies; calculating a comprehensive deviation drought index (CDDI) using a standardized method based on the comprehensive deviation (CD) calculated in step 3; and capturing drought events and analyzing drought characteristics based on the CDDI calculated in step 4. The present invention combines the occurrence of meteorological drought captured by precipitation deviation, considers soil moisture differences that cause agricultural drought development, and incorporates terrestrial water storage deficits to systematically and comprehensively conduct comprehensive drought monitoring.

Description

Technical Field

[0001] The present invention belongs to the technical field of comprehensive drought monitoring, and in particular relates to a drought monitoring method integrating a comprehensive deviation drought index with multiple indicators. Background Art

[0002] Drought is a complex event involving hydrological and atmospheric processes, and is even closely linked to phenological development and socioeconomic development. Drought indices are tools for monitoring, quantifying, and providing early warning of the impact of water scarcity. They quantify one or more characteristic factors into a single numerical value to assess drought. With the passage of time and the maturation of technology, various drought indices have been developed across different fields and disciplines, such as the Meteorological Drought Index (SPI) and the Drought Index (PDSI), the Hydrological Drought Index (SRI) and the Drought Index (SWSI), and the Agricultural Drought Index (SSI) and the Drought Index (SMDI). These indices can be based on site monitoring data, hydrological simulation data, or remote sensing data. In recent years, the long-term, large-scale terrestrial water storage data returned by the GRACE satellites has become a hot topic in drought research. This is due to the advantages of the GRACE gravity satellites, which are independent of ground conditions and can produce stable, uniformly distributed data at a uniform observation scale.

[0003] Terrestrial water storage data from GRACE have been used in practical drought monitoring. Drawing on the iterative principle of the PDSI, some researchers have derived the current TSDI index by combining the previous month's TSDI index with the current water storage deficit. This index was then used to investigate the 2002-2003 drought in Canada. Cao et al. (2015) used the TSDI to capture the spatiotemporal distribution of drought events in northwestern China, confirming the potential of GRACE for regional drought monitoring in China. Numerous studies have also been conducted based on water storage deficits. Wang et al. (2014) analyzed the spatiotemporal variability of the terrestrial water storage index (TWSI), precipitation, and vegetation index (EVI). By removing the mean of the annual variation and deriving the outliers as the corresponding drought index, they analyzed the 2003-2013 drought in the Haihe River Basin of China. However, this approach directly ignores the spatiotemporal variability of drought events. On the other hand, Thomas et al. (2014) proposed the concepts of monthly terrestrial water storage anomalies and monthly climatology, quantified the terrestrial water storage deficit by calculating the deviation between the two, and monitored the drought in parts of the United States between 2003 and 2013. Sinha et al. (2019) proposed the Combined Climate Deviation Index (CCDI) to integrate the deviations of precipitation and terrestrial water storage, and confirmed the validity and applicability of the index through an evaluation of Indian basins. Satish Kumar et al. (2021) analyzed the correlation of five drought indices in four basins in India and found a high correlation between CCDI and GRACE-DSI. They proposed that the combination of indices can better understand drought.

[0004] Although there are many drought indices, the applicability of the indices is limited. Terrestrial water storage can reflect the changes in the quality of terrestrial water storage and determine the final severity of the drought. In addition, precipitation is the only source of moisture for the underlying surface and is a subjective factor causing drought conditions. The high-frequency variability of precipitation can keenly reflect the impact of climate on drought. The CCDI index is a combination of precipitation and TWS. However, after a period of time after the occurrence or end of meteorological drought, it may lead to varying degrees of soil moisture shortage, which in turn triggers the occurrence of agricultural drought ( and 2014). The transmission mechanism between drought and WSDI has attracted the attention of some scholars, but a clear answer remains unanswered (Apurv et al., 2017; Ding et al., 2021). The WSDI primarily reflects deep subsurface moisture anomalies and is insensitive to changes in surface moisture, while the CCDI index does not incorporate agricultural impacts. Therefore, developing a method that incorporates both precipitation and soil moisture series into a comprehensive drought index is both crucial and logically necessary. Summary of the Invention

[0005] The purpose of the present invention is to address the shortcomings of the existing technology and provide a drought monitoring method that integrates a comprehensive deviation drought index with multiple indicators. This method combines the occurrence of meteorological drought captured by precipitation deviation, considers the soil moisture differences that cause the development of agricultural drought, and then adds the land water storage deficit status, so that comprehensive drought monitoring can be carried out systematically and comprehensively.

[0006] In order to solve the above technical problems, the present invention adopts the following technical solutions:

[0007] A drought monitoring method based on a multi-index integrated deviation drought index comprises the following steps:

[0008] Step 1: Calculate the monthly mean climate dataset for each calendar month during the study period by coordinating the data from the terrestrial water storage (TWS), precipitation (PET), and soil moisture (SM) products;

[0009] Step 2: Based on the monthly mean climate dataset obtained in step 1, calculate the monthly climate anomalies of land water storage, precipitation, and soil moisture in the jth month of the i-th year;

[0010] Step 3: normalize the monthly climate anomaly values calculated in step 2, and construct the comprehensive deviation CD based on the normalized climate anomaly values;

[0011] Step 4, using the standardized method to calculate the comprehensive deviation drought index CDDI based on the comprehensive deviation CD calculated in step 3;

[0012] Step 5: Capture drought events and analyze drought characteristics based on the CDDI values calculated in step 4.

[0013] Furthermore, step 1 specifically includes:

[0014] Terrestrial water storage data, precipitation data, and soil moisture data were collected from the data center. The collected data were formatted and the spline interpolation method was used to fill the data gaps for missing months. The processed data were clipped according to the study area, and the long-term average values of terrestrial water storage, precipitation, and soil moisture for each calendar month from January to December were calculated based on the clipped data to obtain the monthly average climate dataset.

[0015] Furthermore, the specific steps are:

[0016] Step 3.1: Standardize the monthly climate anomaly values calculated in step 2. The calculation formulas for the standardization of precipitation deviation, land water storage deviation, and soil moisture deviation are as follows:

[0017]

[0018]

[0019]

[0020] In the formula, PETA i,j PETA μ PETA σ The corresponding monthly climate anomaly, mean precipitation, and standard deviation of precipitation in the jth month of the i-th year;

[0021] TWSA i,j 、TWSA μ 、TWSA σ Correspondingly, it represents the monthly climate anomaly value of land water storage in the jth month of year i, the mean value of land water storage, and the standard deviation of land water storage;

[0022] SMA i,j , SMA μ , SMA σ Corresponding to the monthly climate anomaly value, mean value, and standard deviation of soil moisture in month j of year i;

[0023] CD PETA 、CD TWSA 、CD SMA They correspond to the deviations of precipitation, terrestrial water storage, and soil moisture, respectively;

[0024] Step 3.2: Based on the standardized indicator deviation calculated in step 3.2, the comprehensive deviation is obtained by polynomial combination:

[0025] CD=CD PETA +CD TWSA +CD SMA .

[0026] Furthermore, in step 4, the z-score normalization method is used to calculate the comprehensive deviation drought index CDDI, which is:

[0027]

[0028] Where, CDDI i，j represents the comprehensive deviation drought index of the jth month of the i-th year, CD i，j Denotes the comprehensive deviation CD of the jth month of the i-th year, CD μ 、CD σ They correspond to the mean and standard deviation of the comprehensive deviation respectively.

[0029] Furthermore, step 5 specifically includes:

[0030] Set the drought threshold D0. When the CDDI value calculated in step 4 is lower than D0 for three consecutive months or more, it is considered a drought event.

[0031] With reference to the run theory, drought characteristic quantities are separated according to drought events: drought intensity, drought severity and drought duration. Among them, drought duration is the duration of drought, drought intensity is the sum of CDDI values within the drought duration, and drought severity is the minimum CDDI value within the drought duration.

[0032] Compared with the prior art, the present invention has the following beneficial effects:

[0033] 1. This invention constructs a comprehensive hydrological-meteorological-agricultural drought indicator: This model combines hydrological terrestrial water storage products, meteorological precipitation observations, and agricultural soil moisture data, overcoming the limitations of a single indicator or variable in characterizing drought from a single area of water deficit. Precipitation is the only source of groundwater and directly contributes to drought. However, the high-frequency variability and uneven spatiotemporal distribution of precipitation are characteristics that lead to poor regional adaptability to meteorological drought. Terrestrial water storage reflects the total water storage capacity of the hydrological cycle and can describe the response of ecosystems to changes in water supply. It is robust for drought monitoring at large spatial scales. Soil moisture is very important for agricultural drought caused by the lag effect of precipitation. Therefore, combining precipitation, terrestrial water storage, and soil moisture to develop a comprehensive drought index can reflect the combined effects of environmental factors in multiple fields and realize comprehensive drought monitoring in different fields of meteorology, hydrology, and agriculture.

[0034] 2. The present invention realizes the combination of multiple indicator deviations based on the linear combination method, which is a simple and fast deviation combination method: At present, many researchers at home and abroad have proposed many indicator comprehensive schemes, such as weight combination, multivariate joint sum and its learning, etc. The index construction schemes vary according to the choice of index. After comparing the index construction schemes of PCA and TSNE, the present invention found that the monitoring effect of the simple linear combination scheme is the most stable, and the PN index construction scheme is less restricted by regional restrictions. Whether in the humid south or the dry north, comprehensive drought monitoring can be carried out systematically and comprehensively. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 This is a flow chart of the drought monitoring method based on the combined deviation drought index proposed in an embodiment of the present invention;

[0036] Figure 2 Three methods are used in the present invention to construct the correlation comparison results of CDDI with SPEI, SSWI and SRI in the Chinese basin; among them, (a) the correlation comparison of CDDI and common drought indices constructed based on PCA, TSNE and PN; (b) the box plot statistics of the correlation comparison;

[0037] Figure 3 This is a CDDI time series diagram of the ten major river basins from 2003 to 2020 according to an embodiment of the present invention;

[0038] Figure 4 This is a schematic diagram of the run-length theory used in an embodiment of the present invention. DETAILED DESCRIPTION

[0039] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the embodiments of the present invention. Obviously, the embodiments described are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0040] It should be noted that, in the absence of conflict, the embodiments of the present invention and the features in the embodiments may be combined with each other.

[0041] The present invention will be further described below with reference to specific examples, but they are not intended to limit the present invention.

[0042] like Figure 1 As shown, the present invention discloses a drought monitoring method for a comprehensive deviation drought index integrating multiple indicators, comprising the following steps:

[0043] Step 1: Calculate the monthly mean climate dataset for each calendar month during the study period by coordinating the data from the terrestrial water storage (TWS), precipitation (PET), and soil moisture (SM) products;

[0044] In this example, the terrestrial water storage data, precipitation data, and soil moisture data for the Chinese region between January 2003 and December 2020 were collected from the GRACE, GLDAS, and GPCC data centers, respectively;

[0045] The collected data were then preprocessed and resampled to a 0.25° × 0.25° spatial grid for quality control. For soil moisture data, we aggregated the volumetric water content data of the soil layer together for calculation, i.e., the soil moisture data input was soil moisture data from 0 to 200 cm.

[0046] The GRACE dataset has 20 months of missing data between 2003 and 2020: June 2003, January 2011, June 2011, May 2012, October 2012, March 2013, August 2013, September 2013, February 2014, July 2014, December 2014, June 2015, October 2015, November 2015, April 2016, September 2016, October 2016, February 2017, August 2018, and September 2018. Data for these 20 months were interpolated using cubic spline interpolation using the terrestrial water storage data and trend estimates from the preceding and following months. There was an 11-month gap between the GRACE and GRACE-FO missions (July 2017–May 2018). Because the missing data are adjacent, interpolation errors can occur. Therefore, a dataset reconstructed based on precipitation is used as the estimate of land water storage changes during the gap.

[0047] The data obtained in the above steps are then clipped according to the study area; finally, the long-term average values of terrestrial water storage, precipitation, and soil moisture for each calendar month from January to December are calculated to obtain the monthly average climate dataset;

[0048]

[0049] In the above formula, clim represents the monthly average climate, j represents the month and its value ranges from 1 to 12, and N represents the number of years.

[0050] Step 2: Based on the monthly mean climate dataset obtained in step 1, combined with the time series data of precipitation, terrestrial water storage, and soil moisture in the jth month of the i-th year, calculate the monthly climate anomaly values of terrestrial water storage, precipitation, and soil moisture in the jth month of the i-th year, and use the monthly climate anomaly value as the quantitative standard for the deviation of a specific calendar month from the monthly mean climate normal value; specifically,

[0051]

[0052] In the formula, PETA i，j , TWSA i，j , SMA i，j PET represents the monthly climate anomaly values of precipitation, land water storage and soil moisture in the jth month of the i-th year, respectively. i，j , TWS i，j , SM i，j represent the time series data of precipitation, land water storage and soil moisture in the jth month of the i-th year, respectively.

[0053] Step 3: normalize the monthly climate anomaly values calculated in step 2, and construct the comprehensive deviation CD based on the standardized climate anomaly values;

[0054] In this embodiment, the step specifically includes:

[0055] In step 3.1, the monthly climate anomaly values of the multiple indicators obtained in step 2 are standardized to reduce the impact of drought differences between indicators. The calculation formulas for the standardization of precipitation deviation, land water storage deviation, and soil moisture deviation are as follows:

[0056]

[0057]

[0058]

[0059] In the formula, PETA i,j PETA μ PETA σ The corresponding monthly anomaly, mean precipitation, and standard deviation of precipitation in the jth month of the i-th year;

[0060] TWSA i,j 、TWSA μ 、TWSA σ Correspondingly, it represents the monthly anomaly value of terrestrial water storage, the mean value of terrestrial water storage, and the standard deviation of terrestrial water storage in the jth month of the i-th year;

[0061] SMA i,j , SMA μ , SMA σ Correspondingly, it represents the monthly anomaly value, mean value, and standard deviation of soil moisture in month j of year i;

[0062] CD PETA 、CD TWSA 、CD SMA They correspond to the deviations of precipitation, terrestrial water storage, and soil moisture, respectively;

[0063] Step 3.2: Based on the standardized indicator deviation calculated in step 3.2, perform deviation combination to obtain the comprehensive deviation:

[0064] Deviation combination can be achieved through three methods: PCA, TSNE and PN; among them, polynomial (PN), principal component analysis (PCA) and t-distributed stochastic neighbor embedding (TSNE) are used to integrate the monthly climate deviations respectively. Finally, by comparing the correlation with the Standardized Precipitation Evapotranspiration Index (SPEI), the Standardized Soil Water Index (SSWI) and the Standardized Runoff Index (SRI), it is determined that PN based on the polynomial method has the best effect. Therefore, in this embodiment, a linear combination of the standardized monthly climate deviations of the three indicators based on the polynomial synthesis method is used, that is, CD PN =CD PETA +CD TWSA +CD SMA .

[0065] Step 4, using the standardized method to calculate the comprehensive deviation drought index CDDI based on the comprehensive deviation CD calculated in step 3;

[0066] In this embodiment, the CDDI is obtained using the z-score normalization method, and the comprehensive deviation drought index CDDI is:

[0067]

[0068] Where, CDDI i，j represents the comprehensive deviation drought index of the jth month of the i-th year, CD i，j Denotes the comprehensive deviation CD of the jth month of the i-th year, CD μ 、CD σ They correspond to the mean and standard deviation of the comprehensive deviation respectively.

[0069] Figure 2 The correlation coefficient statistics between the CDDI index constructed based on PCA, TSNA and PN methods and the SPEI, SSWI and SRI drought indices for 10 basins are given. Figure 2(a) In the four basins of NWB, YEB, HRB and LRB in northern China, the correlation between the CDDI_TSNE drought index and the general index is poor. Among the five basins in southern China, PCA shows poor correlation in the three basins of SWB, YZRB and HHRB. For the three basins of SRB, PRB and SEB, there is a good correlation between the three CDDI and the general index. The CDDI index constructed based on PN has a high correlation with the commonly used indices. In comparison with SPEI, the basins with the highest and lowest correlations between CDDI_PN and SPEI are SEB (0.78) and HRB (0.44), respectively. The consistency with SSWI is the highest, with a correlation of 0.81 in the SRB basin, and the worst correlation is in the YRB basin (0.55). According to Figure 2 (b) It is not difficult to find that the correlations between the CDDI indices constructed based on PCA and TSNE are quite different. However, the ability of the CDDI index constructed based on PN to characterize drought in the ten basins is not much different overall, mainly reflected in its relatively stable correlation with the three indices.

[0070] Step 5: Capture drought events and analyze drought characteristics based on the CDDI values calculated in step 4. This step also includes:

[0071] Step 5.1: Set the drought threshold D0. A drought event is identified when the CDDI value is below D0 for three consecutive months or more. In this embodiment, the 30th, 20th, 10th, 5th, and 2nd percentiles of the standard normal distribution can be set as drought thresholds. As the percentile decreases, the drought intensity increases. According to the z-quantile and the standardized normal distribution table, the thresholds corresponding to the five drought levels of CDDI are D0 (-0.5), D1 (-0.8), D2 (-1.2), D3 (-1.6), and D4 (-2.0). Areas without drought are not included in the statistics. Based on drought assessment experience, a drought event is defined as a CDDI value less than -0.5 for three consecutive months or more.

[0072] Figure 3 The captured drought events were numbered and different color bars were drawn according to the drought severity. The darker the color, the greater the drought severity.

[0073] Step 5.2: Figure 4 This is a conceptual diagram of events identified by run theory. When the disaster index falls below the threshold D0 and lasts for more than a certain period (3 months), it is considered a disaster event. Referring to run theory, we separate drought characteristics: drought intensity, drought severity, and drought duration. Drought duration is the duration of drought, drought intensity is the sum of the comprehensive deviation drought index within the drought duration, and drought severity is the minimum comprehensive deviation drought index value within the drought duration, as shown in the following example: Figure 4As shown, according to Figure 4 The drought events captured by CDDI are characterized, and detailed information on the characteristics of drought events in the ten major basins is provided in Appendix Table 1.

[0074] Table 1 Statistics of drought characteristics in China's ten major river basins

[0075]

[0076]

[0077] Based on the calculation results in Table 1 above, a visualization of regional drought characteristics across China from 2003 to 2020 was created. The visualization reveals that Region A in the Yangtze River Basin in southern China has a high frequency of droughts. Over the past 18 years, it has experienced approximately 8-12 droughts. Furthermore, droughts in this region have lasted for a long time, approximately 40-60 months. Drought intensity and severity are high, indicating that these areas have experienced severe droughts. Region B in the Songhua River Basin demonstrates long-term droughts, with individual droughts of relatively mild severity, but overall drought intensity is severe. Region C in the Yangtze River Basin and Region D in the southern Northwest Basin contrast with Region B. These regions have low frequency and duration of droughts, and drought intensity is not as severe as in Region B, but they are concentrated areas of the highest drought intensity. Region E, similar to Region B, has experienced long-term droughts over the past 18 years, but individual droughts have been relatively mild. Region F has experienced less frequent, severe, and intensified droughts, indicating that drought damage in this region is relatively minimal.

[0078] The present invention uses soil moisture products, precipitation products and terrestrial water storage products to construct a comprehensive deviation model, and adopts a simple polynomial combination method to calculate the comprehensive deviation drought index. The proposed method is tested using China's ten major river basins. The results show that the proposed drought monitoring method can capture drought events that have been reported or studied, confirming the effectiveness of the comprehensive deviation drought index.

[0079] The above are only preferred embodiments of the present invention and do not limit the implementation mode and protection scope of the present invention. For those skilled in the art, it should be aware that all solutions obtained by equivalent substitutions and obvious changes made using the contents of the present invention specification should be included in the protection scope of the present invention.

Claims

1. A drought monitoring method integrating a multi-index comprehensive deviation drought index, characterized in that: The steps include: Step 1: Calculate the monthly mean climate dataset for each calendar month during the study period by coordinating the terrestrial water storage (TWS), precipitation (PET), and soil moisture (SM) product data. Step 2: Based on the monthly mean climate dataset obtained in step 1, calculate the monthly climate anomalies of land water storage, precipitation, and soil moisture in the jth month of the i-th year; Step 3: normalize the monthly climate anomaly values calculated in step 2, and construct the comprehensive deviation CD based on the normalized climate anomaly values; Step 4, using the standardized method to calculate the comprehensive deviation drought index CDDI based on the comprehensive deviation CD calculated in step 3; Step 5, capture drought events and analyze drought characteristics based on the CDDI values calculated in step 4; The specific steps are: Step 3.1: Standardize the monthly climate anomaly values calculated in step 2. The calculation formulas for the standardization of precipitation deviation, land water storage deviation, and soil moisture deviation are as follows: ； ； ； Where, 、 、 The corresponding monthly climate anomaly, mean precipitation, and standard deviation of precipitation in the jth month of the i-th year; 、 、 Correspondingly, it represents the monthly climate anomaly value of land water storage in the jth month of year i, the mean value of land water storage, and the standard deviation of land water storage; 、 、 Corresponding to the monthly climate anomaly value, mean value, and standard deviation of soil moisture in month j of year i; 、 、 They correspond to the deviations of precipitation, terrestrial water storage, and soil moisture, respectively; Step 3.2: Based on the standardized indicator deviation calculated in step 3.2, the comprehensive deviation is obtained by polynomial combination: 。 2. The drought monitoring method of the integrated deviation drought index integrating multiple indicators according to claim 1 is characterized in that: Step 1 specifically includes: Terrestrial water storage data, precipitation data, and soil moisture data were collected from the data center. The collected data were formatted and the spline interpolation method was used to fill the data gaps for missing months. The processed data were clipped according to the study area, and the long-term average values of terrestrial water storage, precipitation, and soil moisture for each calendar month from January to December were calculated based on the clipped data to obtain the monthly average climate dataset.

3. The drought monitoring method of the integrated deviation drought index integrating multiple indicators according to claim 1 is characterized in that: In step 4, the z-score normalization method is used to calculate the comprehensive deviation drought index CDDI, which is: ； Where, Indicates the Year The monthly composite deviation drought index, Indicates the Year Monthly comprehensive deviation 、 、 They correspond to the mean and standard deviation of the comprehensive deviation respectively.

4. The drought monitoring method of the integrated deviation drought index integrating multiple indicators according to claim 1 is characterized in that: Step 5 specifically includes: Setting drought thresholds , when the calculated value in step 4 is The value is lower than 3 months or more It is identified as a drought event; With reference to the run theory, drought characteristics are separated according to drought events: drought intensity, drought severity and drought duration. Among them, drought duration is the duration of drought, and drought intensity is the duration of drought. The drought intensity is the smallest during the drought duration. value.

Images