Method for detecting drought response time of vegetation

Abstract

The invention discloses a method for detecting the response time of vegetation to drought, which judges the response time of vegetation to drought based on a standardized rainfall evapotranspiration index (SPEI) and a normalized vegetation index (NDVI), and specifically comprises the following steps: acquiring the normalized vegetation index and the standardized rainfall evapotranspiration index; carrying out downscaling on the standardized precipitation evapotranspiration index product SPEI 1-12 to a certain value; extracting NDVI and SPE I of vegetation, cultivated land, forest land and grassland areas according to the land utilization data; according to a Pearson correlation coefficient method, cumulative effects and time-lag effects of vegetation, forest land, cultivated land and grassland on drought in the Chinese range are analyzed; the method provides a basis for formulating a scientific drought coping strategy.

Description

Technical Field

[0001] The invention relates to the technical field of drought monitoring, in particular to a method for detecting the response time of vegetation to drought. Background Art

[0002] Drought is a persistent water shortage phenomenon caused by regional water balance or imbalance between supply and demand. It is a common climate phenomenon in the world. It occurs frequently, covers a wide range, and lasts for a long time, making it one of the most important meteorological disasters affecting human production and life. The frequent occurrence of drought not only brings huge losses to the national economy, but also to agricultural production. It also causes many adverse ecological impacts such as water shortage, increased sandstorms and desertification. Therefore, the study of drought is urgent. Drought index is an effective tool for studying the evolution characteristics of drought. For example, the Standardized Precipitation Evapotranspiration Index (SPEI) is a drought index proposed based on the analysis of the advantages and disadvantages of PDSI and SPI. It objectively characterizes the comprehensive effect of precipitation and evaporation on drought and is an ideal tool for drought monitoring. In addition, the water deficit caused by drought will affect the normal growth of vegetation, and the growth of vegetation can also reflect the drought conditions in the region. Due to the high correlation between drought and vegetation growth, many studies have investigated the relationship between drought and vegetation growth. For example, Hua et al. used correlation coefficients to study the relationship between NDVI and SPEI and found that drought during the growing season in northern China played a key role in vegetation growth. Vegetation growth is usually not driven by current water and heat conditions, but by earlier drought events. Therefore, determining the timing of vegetation response to drought is crucial to a deeper understanding of the interaction mechanism between vegetation and climate, and to the recognition and development of more effective vegetation protection measures.

[0003] The time lag effect indicates the impact of the previous drought event that occurred at a specific time point on the current vegetation growth, reflecting the drought sensitivity of vegetation. The cumulative effect refers to the impact of a period of water shortage on vegetation growth. Based on the time lag effect, including the dynamic changes of climate conditions (such as precipitation and evaporation) in a specific period, it can be used to evaluate the tolerance of vegetation to drought. Previous researchers have extensively studied the effects of time lag effect and cumulative effect on vegetation growth. Zhan et al. revealed that 40.4% of the vegetation area in the Yellow River Basin was affected by cumulative drought, the accumulation time in the semi-humid area was the longest (eight months), and 91.8% of the vegetation area was affected by lagged drought. Zhao et al. found that vegetation in the arid area of ​​the Loess Plateau had stronger cumulative and time lag effects. Wu et al. reported that the time lag effect of vegetation response was significantly different worldwide. Zuo et al. found that temperature and precipitation in the upper reaches of the Yarlung Zangbo River Basin had a 1-month lag effect on vegetation, temperature in the middle reaches had a 1-month lag effect, and temperature and precipitation in the lower reaches had no lag effect on vegetation. Hu et al. found that soil moisture and climate factors in the northwest of the Yangtze River Basin had a one-month lag. Ji et al. revealed that vegetation in the Great Plains of the United States has a 3-month time lag to precipitation during the growing season. Hua et al. found that vegetation in Nebraska, the United States, has a 30-45 day time lag to drought.

[0004] Specifically, the purpose of studying the cumulative effect and time lag effect of drought is to reveal the impact mechanism of drought events on water cycle, soil moisture, vegetation growth, etc., so as to better understand the occurrence and development of drought events. In addition, this kind of research can also better evaluate the impact range and duration of drought events, and improve the accuracy and timeliness of drought monitoring and prediction. At the same time, studying the cumulative effect and time lag effect of drought can also reveal the impact of drought events on water resources, ecological environment and social economy, and provide a basis for formulating scientific drought response strategies. This kind of research is helpful to promote sustainable development, because drought is one of the important factors affecting sustainable development. By studying the cumulative effect and time lag effect of drought, we can better evaluate the impact of drought events on sustainable development and provide guidance for achieving sustainable development. Summary of the invention

[0005] In order to solve the problems existing in the prior art, the purpose of the present invention is to provide a method for detecting the response time of vegetation to drought, and the present invention provides a basis for formulating scientific drought response strategies.

[0006] To achieve the above object, the technical solution adopted by the present invention is: a method for detecting the response time of vegetation to drought, judging the response time of vegetation to drought based on the standardized precipitation evapotranspiration index SPEI and the normalized vegetation index NDVI, specifically comprising the following steps:

[0007] Step 1, obtaining the normalized vegetation index and the standardized precipitation evapotranspiration index;

[0008] Step 2: downscaling the standardized precipitation evapotranspiration index product SPEI1-12 to a certain value;

[0009] Step 3: Extract NDVI and SPEI of vegetation, cultivated land, forest land and grassland areas based on land use data;

[0010] Step 4: Analyze the cumulative effects and time lag effects of vegetation, forest land, cultivated land, and grassland on drought in China based on the Pearson correlation coefficient method.

[0011] As a further improvement of the present invention, in step 1, the Normalized Difference Vegetation Index NDVI is extracted from MODIS data using the ModisToolReprojection tool, and the NDVI and SPEI data are uniformly projected using ArcPy.

[0012] As a further improvement of the present invention, in step 2, it also includes: using ArcPy to batch-crop the NDVI and SPEI data to obtain data within a certain range, and generating a raster data file in TIFF format.

[0013] As a further improvement of the present invention, in step 2, the standardized precipitation evapotranspiration index product SPEI1-12 is downscaled to 1 km.

[0014] As a further improvement of the present invention, in step 4, the time scale and time range of SPEI are considered according to the Pearson correlation coefficient analysis of the cumulative effect and the time lag effect. The cumulative effect uses SPEI1-12 with a cumulative scale of consecutive years to perform correlation analysis with NDVI respectively, and the time lag effect uses SPEI1 with a time scale of 1 month for consecutive years to perform correlation analysis with NDVI for consecutive years.

[0015] As a further improvement of the present invention, in step 4, the correlation between the vegetation index NDVI and the drought index SPEI at each cumulative scale is analyzed by the following formula:

[0016] r j =corr(NDVI,SPEI j )1≤j≤12

[0017] The following formula is used to analyze at which cumulative scale the correlation between NDVI and SPEI is strongest:

[0018] r max-cul =max(r j )1≤j≤12

[0019] In the formula, r j represents the pixel-by-pixel correlation coefficient between NDVI and SPEI at the jth time scale, where j is the cumulative time of 1-12 months, and r max_culYes j The maximum value of .

[0020] The beneficial effects of the present invention are:

[0021] The correlation coefficient is a simple statistical tool with an easy-to-understand calculation method. It can quickly obtain the cumulative effect and time lag effect between the drought index and the vegetation index. It is very useful to quickly understand the impact of drought events on the ecosystem and socio-economy in a short period of time. It is very important for formulating management strategies for drought events, planning drought prevention and control work, and predicting the impact range of drought. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 is a flow chart of an embodiment of the present invention. DETAILED DESCRIPTION

[0023] The embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0024] Example 1

[0025] like Figure 1 As shown, a method for detecting the response time of vegetation to drought events comprises the following steps:

[0026] (1) Obtain the Normalized Difference Vegetation Index and Standardized Precipitation Evapotranspiration Index;

[0027] (2) Downscaling the Standardized Precipitation Evapotranspiration Index products SPEI 1-12 to 1 km;

[0028] (3) Analyze the correlation between vegetation index NDVI and drought index SPEI at each cumulative scale using formula (1);

[0029] (4) Using formula (2), we analyzed at which cumulative scale the correlation between NDVI and SPEI was strongest;

[0030] r j =corr(NDVI,SPEI j )1≤j≤12 (1)

[0031] r max-cul =max(r j )1≤j≤12 (2)

[0032] In the formula, r j represents the pixel-by-pixel correlation coefficient between NDVI and SPEI at the jth time scale, where j is the cumulative time of 1-12 months, and r max_cul Yes j The maximum value of .

[0033] (5) Use land use data to extract NDVI and SPEI for cultivated land, forest land, and grassland areas;

[0034] (6) Repeat steps (3) and (4) to analyze the cumulative number of months for cultivated land, forest land, and grassland in different regions;

[0035] (7) Analyze the correlation between vegetation index NDVI and drought index SPEI at various time lag scales using formula (3);

[0036] (8) Formula (4) is used to analyze at which time lag scale the correlation between NDVI and SPEI is strongest.

[0037] r i =corr(NDVI,SPEI i )1≤i≤12 (3)

[0038] r max-lag =max(r i )0≤i≤12 (4)

[0039] In the formula, r i represents the pixel-by-pixel correlation coefficient between NDVI and SPEI at the i-th time scale, i is the time lag of 0-12 months (0 means no time lag, 1-12 means time lag of 1-12 months), r max_lag Yes i The maximum value of .

[0040] (9) Repeat steps (7) and (8) to analyze the time lag in months for cultivated land, forest land, and grassland in different regions.

[0041] Example 2

[0042] A method for detecting the response time of vegetation to drought events. This embodiment uses MODIS products and SPEI in China from 2001 to 2020 as data sources to monitor the accumulation and time lag effects of drought in China. Figure 1 As shown, the specific implementation steps are as follows:

[0043] Step 1: Obtain the monthly 1km dataset of MOD13A3 from NASA’s official website, use Modis ToolReprojection (MRT) to batch extract NDVI bands, mosaic, and crop the study area to obtain 240 monthly NDVI images of China.

[0044] Step 2: Resample the SPEI1 data from 2000 to 2020 and the SPEI2-12 data from 2001 to 2020 to 1 km.

[0045] Step 3: Obtain the land use data for 2005 and 2020, use the raster calculator to obtain the raster layer in which the land use pattern has not changed between 2005 and 2020, and extract the raster data of cultivated land, forest land, and grassland.

[0046] Step 4: Calculate the cumulative effect.

[0047] Formula (1) was used to perform correlation analysis on NDVI and SPEI1-12 from 2001 to 2020, and 12 correlation coefficient raster layers were obtained;

[0048] The maximum correlation coefficient of each pixel in the 12 layers was analyzed by using the pixel statistics tool in the ArcGIS toolbox through formula (2) to obtain the maximum correlation coefficient spatial distribution map of the cumulative effect;

[0049] r j =corr(NDVI,SPEI j )1≤j≤12 (1)

[0050] r max-cul =max(r j )1≤j≤12 (2)

[0051] Among them, r j represents the pixel-by-pixel correlation coefficient between NDVI and SPEI at the jth time scale, where j is the cumulative time of 1-12 months, and r max_cul Yes j The maximum value of .

[0052] The highest position tool was used to obtain the cumulative number of months corresponding to the maximum correlation coefficient of each pixel, and the spatial distribution map of the cumulative effect was obtained.

[0053] The calculation steps for the cumulative effects of cultivated land, forest land and grassland are the same as above.

[0054] Step 5: Calculate the time lag effect.

[0055] Formula (3) is used to perform correlation analysis on the NDVI from 2001 to 2020 and the SPEI 1 from 2000 to 2020. It should be noted that the time lag of 0 months is the correlation analysis of the NDVI and SPEI 1 from 2001 to 2020, and the time lag of 1 month is the correlation analysis of the NDVI from 2001 to 2020 and the SPEI 1 from December 2000 to November 2020, and so on, to obtain 13 correlation coefficient raster layers;

[0056] Similar to step 4, the maximum correlation coefficient of each pixel in the 13 layers is analyzed by superimposing the pixel statistics tool through formula (4) to obtain the spatial distribution map of the maximum correlation coefficient of the time lag effect;

[0057] r i =corr(NDVI,SPEI i )1≤i≤12 (3)

[0058] r max-lag =max(r i )0≤i≤12 (4)

[0059] Among them, r i represents the pixel-by-pixel correlation coefficient between NDVI and SPEI at the i-th time scale, i is the time lag (0 means no time lag, 1 to 12 means a time lag of 1 to 12 months), r max_lag Yes i The maximum value of .

[0060] Using the highest position tool, we can obtain the time lag months corresponding to the maximum correlation coefficient of each pixel and obtain the spatial distribution map of the time lag effect. The calculation steps for the time lag effect of cultivated land, forest land and grassland are the same as above.

[0061] Cumulative and time-lag effect results:

[0062] According to the spatial distribution map of the cumulative effect of the whole country, cultivated land, forest land and grassland. The best cumulative effect shows obvious regionality. The cumulative effect in Northeast China is mainly based on the scale of 9 months, the Northwest is mainly based on the scale of 9-12 months, and the southern Xinjiang shows a large difference, mainly based on the scale of 1-5 months. The Qinghai-Tibet region is mainly based on the scale of 4-7 months. The three provinces of Yunnan, Guizhou and Sichuan are mainly based on the scale of 5-9 months. The scale of North China and Central China is mainly based on the scale of 1-3 months. The scale of South China is 12 months except for the southern coastal areas, and the scale of the southern coastal areas is 1-5 months. From south to north, there is a general trend of increasing cumulative months. Among them, the cumulative effect of 9 months is the most widely distributed (24.58%), followed by 12 months (20.05%).

[0063] According to the spatial distribution map of the time lag effect of the whole country, cultivated land, forest land and grassland. The optimal time lag time is strongly regional in space. The time lag time in Northeast China is mainly 6-8 months, the Loess Plateau is mainly 7-9 months, the northern Xinjiang is 5-7 months, and the time lag time in southern Xinjiang is quite different. The Qinghai-Tibet region is 3-6 months, the three provinces of Yunnan, Guizhou and Sichuan are 2-6 months, the central China region is 8-12 months, the North China region is mainly 8 months, and the South China region is 9-11 months. The time lag time interval is mainly 6-8 months (44.85%).

[0064] Example 3

[0065] A method for detecting the response time of vegetation to drought events, the method steps are as follows:

[0066] Step 1: Obtain MODIS data and standardized precipitation evapotranspiration index data on a 1-12 month scale. Use the Modis Tool Reprojection (MRT) tool to extract the Normalized Difference Vegetation Index (NDVI) band from the MODIS data, and use ArcPy to unify the projection of NDVI and SPEI data.

[0067] Step 2: Downscale the standardized precipitation evapotranspiration index products SPEI1-12 after unified projection to 1 km, and use ArcPy to batch crop the NDVI and SPEI data for the Chinese area to generate raster data files in TIFF format;

[0068] Step 3: Crop the NDVI and SPEI data of forest land, cultivated land and grassland at the 1-12 month scale based on the land use data

[0069] Step 4: Analyze the cumulative effects and time lag effects of vegetation, forest land, cultivated land, and grassland on drought in China based on the Pearson correlation coefficient method.

[0070] In step 4, when analyzing the cumulative effect and lag effect based on the Pearson correlation coefficient, it is necessary to pay attention to the time scale and time range of SPEI. The cumulative effect uses SPEI1-12 with a cumulative scale from 2001 to 2020 for correlation analysis with NDVI respectively, and the lag effect uses SPEI1 with a time scale of 1 month from 2000 to 2020 for correlation analysis with NDVI from 2001 to 2020.

[0071] The above-mentioned embodiments only express the specific implementation of the present invention, and the description thereof is relatively specific and detailed, but it cannot be understood as limiting the scope of the present invention. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, which all belong to the protection scope of the present invention.

Claims

1. A method for detecting the response time of vegetation to drought, characterized in that: The vegetation response time to drought is judged based on the standardized precipitation evapotranspiration index (SPEI) and the normalized difference vegetation index (NDVI), which specifically includes the following steps: Step 1, obtaining the normalized vegetation index and the standardized precipitation evapotranspiration index; Step 2: downscaling the standardized precipitation evapotranspiration index product SPEI 1-12 to a certain value; Step 3: Extract NDVI and SPEI of vegetation, cultivated land, forest land and grassland areas based on land use data; Step 4: Analyze the cumulative effects and time lag effects of vegetation, forest land, cultivated land, and grassland on drought in China based on the Pearson correlation coefficient method.

2. The method for detecting the response time of vegetation to drought according to claim 1, characterized in that: In step 1, the Normalized Difference Vegetation Index (NDVI) is extracted from MODIS data using the Modis Tool Reprojection tool, and the NDVI and SPEI data are uniformly projected using ArcPy.

3. The method for detecting the response time of vegetation to drought according to claim 1, characterized in that: In step 2, it also includes: using ArcPy to batch crop a certain range of data from the NDVI and SPEI data and generate a raster data file in TIFF format.

4. The method for detecting the response time of vegetation to drought according to claim 1, characterized in that: In step 2, the Standardized Precipitation Evapotranspiration Index products SPEI 1-12 are downscaled to 1 km.

5. The method for detecting the response time of vegetation to drought according to claim 1, characterized in that: In step 4, the cumulative effect and time lag effect were analyzed based on the Pearson correlation coefficient, taking into account the time scale and time range of SPEI. The cumulative effect was correlated with NDVI using SPEI 1-12 with a cumulative scale of consecutive years, and the time lag effect was correlated with SPEI 1 with a time scale of 1 month and NDVI for consecutive years.

6. The method for detecting the response time of vegetation to drought according to claim 5, characterized in that: In step 4, the correlation between vegetation index NDVI and drought index SPEI at each cumulative scale is analyzed by the following formula: r j =corr(NDVI,SPEI j )1≤j≤12 The following formula is used to analyze at which cumulative scale the correlation between NDVI and SPEI is strongest: r max-cul =max(r j )1≤j≤12 In the formula, r j represents the pixel-by-pixel correlation coefficient between NDVI and SPEI at the jth time scale, where j is the cumulative time of 1-12 months, and r max_cul Yes j The maximum value of .

Images