CN-122024440-A - Seasonal river water ecological risk early warning method based on environment monitoring instrument data

CN122024440ACN 122024440 ACN122024440 ACN 122024440ACN-122024440-A

Abstract

The invention belongs to the technical field of environmental monitoring, and particularly relates to a seasonal river water ecological risk early warning method based on environmental monitoring instrument data. The method comprises the steps of 1, arranging ultraviolet-visible spectrum probes on monitoring sections of a target seasonal river to form spectrum feature vectors of all the monitoring sections, 2, constructing a maximum entropy habitat model by taking the spectrum feature vectors as environment variables to obtain an aquatic organism fitness field, 3, dividing hydrologic seasons according to hydrologic features of the target seasonal river, setting ecological risk level judging thresholds for different hydrologic seasons, determining ecological risk levels, and generating and pushing early warning information when the ecological risk levels reach preset trigger levels. The invention realizes depth information mining of spectrum data, quantitative characterization of upstream and downstream pollution transmission relation and differential response of seasonal hydrologic characteristics, and improves accuracy, timeliness and interpretability of seasonal river water ecological risk early warning.

Inventors

JIANG NA
CHEN XIUNA
ZHANG MIN
QIAN YUTING

Assignees

山东省聊城生态环境监测中心

Dates

Publication Date: 20260512
Application Date: 20260130

Claims (10)

1. The seasonal river water ecological risk early warning method based on the environmental monitoring instrument data is characterized by comprising the following steps of: Step 1, arranging an ultraviolet visible spectrum probe on a monitoring section of a target seasonal river, collecting an original absorbance spectrum of a water body, preprocessing the original absorbance spectrum, and extracting spectral characteristic parameters to form spectral characteristic vectors of all the monitoring sections; Step 2, constructing a maximum entropy habitat model by taking the spectral feature vector as an environment variable to generate a maximum entropy habitat suitability initial field, calculating spectral transfer entropy for the spectral feature parameter time sequence of each monitoring section to form a spectral transfer entropy matrix, and fusing the maximum entropy habitat suitability initial field with the spectral transfer entropy matrix to obtain an aquatic organism suitability field; And 3, dividing hydrologic seasons according to hydrologic features of the target seasonal river, setting ecological risk level judging thresholds according to different hydrologic seasons, comparing the fusion fitness value of each monitoring section in the aquatic organism fitness field with the ecological risk level judging threshold corresponding to the current hydrologic season, determining an ecological risk level, and generating and pushing early warning information when the ecological risk level reaches a preset trigger level.
2. The method of claim 1, wherein the ultraviolet visible spectrum probe comprises a deuterium lamp light source, a tungsten lamp light source, an optical fiber collimator, a quartz cuvette, a concave grating and a linear array charge-coupled device detector, wherein the deuterium lamp light source emits ultraviolet light with a wavelength range of 200 nm to 400 nm, the tungsten lamp light source emits visible light with a wavelength range of 400 nm to 800 nm, the two paths of light are combined by the optical fiber collimator and then pass through a water sample in the quartz cuvette, and the transmitted light is dispersed by the concave grating and then projected to the linear array charge-coupled device detector to obtain an original absorbance spectrum covering a wavelength range of 200 nm to 800 nm.
3. The method according to claim 1, wherein the preprocessing of the original absorbance spectrum comprises the steps of firstly collecting a dark current signal when an ultraviolet visible spectrum probe is not contacted with a water sample, deducting the dark current signal from the original absorbance spectrum from wavelength points to obtain a dark current correction spectrum, then taking an average value of the dark current correction spectrum in a wave band of 200-230 nanometers and a wave band of 780-800 nanometers as two end datum points respectively, taking a connecting line of the two end datum points as a base line and deducting the connecting line from the dark current correction spectrum to obtain a base line correction spectrum, and smoothing the base line correction spectrum by adopting a Savez-Golay filter to obtain a smooth spectrum.
4. The method of claim 3, wherein the spectral characteristics extracted from the smoothed spectrum include an absorbance value at 254 nm, an absorbance value at 280 nm, a ratio of absorbance value at 254 nm to absorbance value at 365 nm, a 400 nm to 450 nm band absorbance integration area, a spectral slope of 600 nm to 700 nm band, and a ratio of 200 nm to 400 nm band absorbance integration area to 400 nm to 800 nm band absorbance integration area.
5. The method according to claim 1, wherein the construction process of the maximum entropy habitat model comprises the steps of regarding each monitoring section as 1 space grid point, marking the monitoring section where the species existence point is located as an existence grid point, marking the monitoring section where the species existence point is not covered as a background grid point, assigning an initial existence probability to each space grid point, setting the initial existence probability as the reciprocal of the total number of all space grid points, calculating the arithmetic average value of the corresponding item spectrum characteristic parameters on all the existence grid points as a existence constraint value for each spectrum characteristic parameter, calculating the arithmetic average value of the corresponding item spectrum characteristic parameters on all the space grid points as a background expected value, adding 1 adjustment step when the absolute value of the difference value of the corresponding item spectrum characteristic parameter value and the existence constraint value on a certain space grid point is smaller than the absolute value of the difference value of the corresponding item spectrum characteristic parameter value and the background expected value, and normalizing the existence probability of all the space grid points after the probability adjustment of all the spectrum characteristic parameters is completed, and obtaining the maximum entropy habitat fitness value of each monitoring section.
6. The method according to claim 1, wherein the calculation process of the spectral shift entropy comprises the steps of forming a source time sequence by continuous time values of a certain spectral characteristic parameter of an upstream monitoring section, forming a target time sequence by continuous time values of the same spectral characteristic parameter of a downstream monitoring section, respectively performing symbolizing processing on the source time sequence and the target time sequence, coding a value which is larger than the median of a current sequence in the sequence as a sign A, and coding a value which is smaller than or equal to the median of the current sequence as a sign B for each sequence, so as to obtain a symbolized source sequence and a symbolized target sequence.
7. The method of claim 6 wherein after setting the embedding dimension, constructing a sliding window for the symbolized target sequence, counting all possible symbol combination patterns and occurrence frequencies thereof in the symbolized target sequence, pairing the symbolized source sequence and the symbolized target sequence time by time to form a joint symbol pair, counting all possible pairing patterns and joint occurrence frequencies thereof in the joint symbol pair, calculating joint entropy according to the joint occurrence frequencies, calculating target entropy according to the symbol combination pattern occurrence frequencies of the symbolized target sequence, and subtracting the target entropy from the joint entropy to obtain the spectral migration entropy value of the corresponding item spectral characteristic parameter.
8. The method of claim 1, wherein the process of fusing the maximum entropy habitat suitability initial field with the spectral shift entropy matrix comprises, for each monitored section, extracting a spectral shift entropy vector between the current monitored section and an adjacent upstream section, arithmetically averaging each spectral shift entropy value in the spectral shift entropy vector to obtain an average spectral shift entropy value of the current monitored section, and multiplying the maximum entropy habitat suitability value of the current monitored section by a ratio of the average spectral shift entropy value of the current monitored section to a maximum value in the average spectral shift entropy values of all the monitored sections to obtain a fused suitability value of the current monitored section.
9. The method of claim 1, wherein the hydrologic season is divided into a dead water period when the real-time flow rate of the monitored section is less than 50% of the annual average flow rate of the current monitored section, a flat water period when the real-time flow rate of the monitored section is greater than or equal to 50% of the annual average flow rate of the current monitored section and less than or equal to 150% of the annual average flow rate of the current monitored section, and a rich water period when the real-time flow rate of the monitored section is greater than 150% of the annual average flow rate of the current monitored section.
10. The method of claim 1, wherein the early warning information comprises an early warning section number, a current hydrologic season type, a current fusion suitability value, a current ecological risk level, a dominant impact factor number and an upstream transmission intensity index, wherein the dominant impact factor number is a number corresponding to a spectral characteristic parameter with the largest value in a spectral transfer entropy vector of a monitoring section triggering early warning, and the upstream transmission intensity index is a spectral transfer entropy value between the monitoring section triggering early warning and an upstream adjacent section thereof.

Description

Seasonal river water ecological risk early warning method based on environment monitoring instrument data Technical Field The invention belongs to the technical field of environmental monitoring, relates to environmental monitoring through big data analysis, and in particular relates to a seasonal river water ecological risk early warning method based on environmental monitoring instrument data. Background Traditional river water quality monitoring mainly relies on manual sampling and laboratory analysis methods. Monitoring personnel periodically go to a monitoring section to collect a water sample, and after the water sample is sent to a laboratory, various water quality indexes are measured by adopting analysis means such as chemical titration, spectrophotometry, atomic absorption spectrometry and the like. Although the monitoring mode has higher analysis precision, the monitoring mode has obvious problem of insufficient timeliness. The time from sampling to obtaining an analysis result usually needs hours or even days, so that the dynamic change of water quality is difficult to reflect in time, and the quick response requirement of an emergency can not be met. Meanwhile, the frequency of manual sampling is limited by manpower and material resources, and regular monitoring can only be realized once a week or a month, a large amount of water quality fluctuation information is missed during the sampling interval, and the peak time and the change trend of the pollutant concentration cannot be accurately captured. In order to overcome the limitations of the traditional monitoring method, the automatic on-line monitoring technology is widely applied. The on-line monitoring system can realize continuous automatic measurement of conventional parameters such as water temperature, dissolved oxygen, conductivity, turbidity, ammonia nitrogen and the like through a sensor array arranged on a monitoring section, and the data acquisition frequency can reach a minute level. The time resolution of data acquisition is greatly improved by on-line monitoring, and a foundation is laid for fine characterization of dynamic change of water quality. However, the existing online monitoring system adopts a single sensor to measure each parameter respectively, the sensors are mutually independent, and the comprehensive optical characteristics and the organic matter composition characteristics of the water body are difficult to be characterized on the whole. In addition, the measurement principle of the single-parameter sensor is different, the calibration maintenance period is different, and the problems of data drift and inconsistency easily occur in long-term operation. Disclosure of Invention The invention mainly aims to provide the seasonal river water ecological risk early warning method based on the environmental monitoring instrument data, so that the depth information mining of spectrum data, the quantitative characterization of upstream and downstream pollution transmission relations and the differential response of the seasonal hydrologic characteristics are realized, and the accuracy, timeliness and interpretability of the seasonal river water ecological risk early warning are improved. In order to solve the technical problems, the invention provides a seasonal river water ecological risk early warning method based on environmental monitoring instrument data, which comprises the following steps: Step 1, arranging an ultraviolet visible spectrum probe on a monitoring section of a target seasonal river, collecting an original absorbance spectrum of a water body, preprocessing the original absorbance spectrum, and extracting spectral characteristic parameters to form spectral characteristic vectors of all the monitoring sections; Step 2, constructing a maximum entropy habitat model by taking the spectral feature vector as an environment variable to generate a maximum entropy habitat suitability initial field, calculating spectral transfer entropy for the spectral feature parameter time sequence of each monitoring section to form a spectral transfer entropy matrix, and fusing the maximum entropy habitat suitability initial field with the spectral transfer entropy matrix to obtain an aquatic organism suitability field; And 3, dividing hydrologic seasons according to hydrologic features of the target seasonal river, setting ecological risk level judging thresholds according to different hydrologic seasons, comparing the fusion fitness value of each monitoring section in the aquatic organism fitness field with the ecological risk level judging threshold corresponding to the current hydrologic season, determining an ecological risk level, and generating and pushing early warning information when the ecological risk level reaches a preset trigger level. Further, the ultraviolet visible spectrum probe comprises a deuterium lamp light source, a tungsten lamp light source, an optical fiber collimator, a quartz cuvette, a concave gra