Font Size: a A A

Water Chlorophyll A Inversion Model Based On Hyperspectral Data

Posted on:2019-12-04Degree:MasterType:Thesis
Country:ChinaCandidate:Y X HeFull Text:PDF
GTID:2381330623963173Subject:Mathematics
Abstract/Summary:PDF Full Text Request
After entering the 20 th century,with the rapid growth of China's population,the continuous improvement of the level of urban and rural areas,and the continuous increase in the total size of the economy,the level of people's lives and daily needs are increasing,resulting in both the quantity and the size of the enterprise.Increasingly,a large amount of industrial and domestic sewage has been discharged into rivers,lakes,and other bodies of water without any purification treatment or improper treatment in the absence of supervision,which has caused problems such as water pollution and imbalance in water environment carrying capacity in China.It has become a major obstacle to restricting the country's economic development and people's drinking water and physical health.Therefore,under the background of promoting the construction of ecological civilization and advocating a sustainable development economy by the party and the country,it is particularly important to grasp the water quality of the region in relation to the formulation of relevant policies for the region.The water quality of the water body needs to obtain accurate,periodic and time-sensitive monitoring to obtain relevant data to understand the water environment and its changes in the area.Chlorophyll is an important indicator in water quality parameters.Algae and other plants contain chlorophyll a(Chl-a).The concentration of chlorophyll a is closely related to the type and quantity of algae and other plants in the water.Therefore,monitoring the concentration of chlorophyll a in water is also one of the important parameters indicating that the entire water body is in the degree of eutrophication.When the concentration of chlorophyll a in the water is greater than 10 ?g/L,the water is considered to be eutrophic.At present,China's water quality monitoring system is mainly based on the traditional sampling to laboratory analysis.Although this traditional method has high monitoring accuracy it has a low input-output ratio and is constrained by various uncontrollable factors such as the regional environment and weather conditions,making it difficult to quickly obtain large-scale and periodic monitoring data.Hyperspectral inversion technology is faster and cheaper than traditional methods,and has the advantages of time-effectiveness and periodicity.It can make early detection and early detection of regional water quality eutrophication monitoring.Indicates the reference role.It is an important supplement to routine water quality monitoring and provides new technical means for regional water environment monitoring.This paper will use the Lake Lake Reservoir,the Shuangfeng Reservoir and the Gongjiayu Reservoir under the jurisdiction of Pengshan District of Meishan City as the research and application demonstration area,based on the field collected water hyperspectral data and the chemical analysis and inspection data obtained by the traditional method.The water hyperspectral curve collected in the field and the chlorophyll a concentration of the water sample detected by chemical analysis were modeled,and an inversion model of the chlorophyll a concentration can be quickly obtained by using the hyperspectral curve of the water body.The main conclusions of this study are as follows:(1)The treatment of water hyperspectral data in this paper is one of the important tasks for the inversion of water chlorophyll a(Chl-a)concentration(?g/L)through the reflectance information of hyperspectral data.Since the collected hyperspectral data has a lot of noise,it affects the quality of hyperspectral data,and it also restricts the accuracy of the chlorophyll plate model established later.Therefore,based on the characteristics of measured water hyperspectral data,this paper establishes a new wavelet threshold denoising method based on the wavelet threshold denoising method proposed by previous researchers and by selecting appropriate threshold and threshold quantization function.By comparing the traditional hard threshold and soft threshold denoising methods of this new wavelet threshold denoising method,the new wavelet threshold denoising method not only simplifies the calculation,but also has obvious mathematical and physical significance.Through this denoising simulation experiment,we can also find that using this denoising method can also obtain a higher signal-to-noise ratio,and the error is also smaller,and the denoising effect is better.The application of this new wavelet threshold denoising method in real-time hyperspectral data processing can effectively improve the image and signal processing effects,achieve the purpose of improving the signal-to-noise ratio,and highlight the information contained in the reflectance of hyperspectral data.(2)By improving the PROSPECT model of predecessors,it can better simulate the correlation of chlorophyll a in the water environment.With the improved PROPECT model,the hyperspectral curve of the chlorophyll-a reflectivity simulated by the SAIL model was used to screen the sensitivity bands of the hyperspectral curve,which provided the data foundation for the establishment of the water-chlorophyll a inversion model.At the same time,this paper established the HWCI index for hyperspectral water chlorophyll a by referring to the MTCI index.By combining the improved PROSPECT-SAIL plate model with the HWCI index,the reflectance information for the specific "red-rim region" sensitivity band contained in the water hyperspectral curve and the concentration of chlorophyll a in the water measured using traditional chemical analysis methods were established.A quantitative inversion model is the water-chlorophyll a plate model.In addition,the water-chlorophyll-a simulation experiment was conducted in the laboratory and error analysis was performed on the simulation experiment to prove the feasibility of the model.(3)After confirming the feasibility of the water-chlorophyll a plate model,this paper presents an application demonstration for this model in Pengshan District of Meishan City.A 16-group water body hyperspectral data collected from Pengshan District and its corresponding chlorophyll-a concentration obtained by using traditional chemical analysis methods were used to establish Pengshan District water-chlorophyll-a plate model,and the model was subjected to error analysis.The accuracy of the model.After that,the model was brought into the model with data from five other sampling points.The inversion chlorophyll a concentration obtained using the model was compared with the true chlorophyll a concentration obtained by using traditional chemical analysis methods,and the reliability of the model was examined.Finally,this paper combines the assessment results of the water resources that were approved by the experts in the Pilot Project of Evaluation of Resources and Environmental Carrying Capacity in Typical Areas of Chengdu-Chongqing Economic Zone and the water spectrum data collected in this paper through Pengshan District Water-Chlorophyll a plate model.The degree of eutrophication reflected by the concentration of chlorophyll a was compared,which proved the rationality of the model in practical application.
Keywords/Search Tags:Water quantity, chlorophyll-a, hyper-spectral, mathematical model
PDF Full Text Request
Related items