Inversion model of surface bare soil temperature and water content based on UAV thermal infrared remote sensing

https://doi.org/10.1016/j.infrared.2022.104289Get rights and content

Highlights

  • Useful for monitoring regional water content with the goal of landslide warning etc.

  • Enable widescale 3 water content monitoring with high spatial and temporal resolution.

  • Rapid and accurate inversion data of bare surface moisture content.

Abstract

Soil moisture content is important to agricultural monitoring, ecological restoration and landslide warning. Traditional methods cannot monitor soil moisture content in large area, high precision and continuous data. Unmanned aerial vehicle (UAV) remote sensing monitoring has enabled wide-scale water content monitoring, improving the spatial resolution accuracy to centimeter level and the temporal resolution accuracy to minute level, which makes UAV remote sensing favorable for soil moisture content monitoring. In this paper, to achieve rapid and accurate inversion data of bare surface soil moisture content via UAV thermal infrared remote sensing, UAV is used to obtain a visible light map of the experiment area and a thermal infrared radiation map at 9:00 am, 1:00 pm, and 6:00 pm in a given day. Gray value of thermal infrared radiation is based on Planck’s blackbody radiation law and quadratic polynomial fitting to invert temperature data; the ground soil temperature and humidity sensor provides sample areas of the measured temperature and soil moisture content. The linear relationship between gray value of thermal infrared radiation, brightness temperature, measured temperature, and soil moisture content was obtained by analyzing the data. The experimental results show that there is a good linear relationship between the brightness temperature and water content, and the brightness t temperature R2 at 6 PM is better than the measured one. Finally, a model for calculating the surface bare soil moisture content based on brightness temperature is established. RMSE and RPD were used to verify the accuracy of the inversion model. The model accuracy verification results show that the inversion model at 6 pm has a good accuracy, which verifies the feasibility and accuracy of unmanned aerial vehicle thermal infrared remote sensing monitoring of surface bare soil moisture content.

Introduction

Soil moisture content is one of the important parameters of agriculture, ecology, meteorology, geology and other disciplines, which plays an important role in water-saving irrigation, geological disasters, ecological restoration, and agricultural monitoring [1], [2], [3]. In the traditional soil moisture monitoring method, the frequency domain reflection analysis method is based on the point scale, and the calibration experiment is carried out in the laboratory to make the measurement results as close as possible to the real value [4]. Neutron method cannot be used to determine the soil moisture content in the surface layer. Drying method, gamma ray attenuation method and thermal pulse technology can be used to determine the soil moisture content in the surface layer, which has the problem of high cost and unable to achieve large-scale and efficient monitoring [5], [6]. The cosmic ray neutron method has been proved to be able to continuously and accurately measure the average soil moisture content at a certain depth within hundreds of meters, but cannot obtain the continuous soil moisture content in a large area [7] (see Fig. 1).

The traditional monitoring methods cannot meet the current application demands for the production of large-area, high-precision and continuous data of soil moisture products. Remote sensing technology has the advantages of rapid acquisition, large-area and real-time, which provides convenient conditions for predicting the spatial and temporal information of soil moisture in large areas and real-time monitoring the dynamic change of soil moisture [8].

In terms of satellite remote sensing, the brightness temperature observations of multiple satellites on the same day jointly estimate the soil moisture, and the average soil moisture of each satellite is retrieved to obtain the composite daily soil moisture product with a spatial resolution of 25 km × 25 km [9]. Based on the multi-temporal synthetic aperture radar (SAR) data, the soil moisture content in bare land was estimated with a time resolution of one month [10]. SAR and optical images are fused to retrieve the soil moisture of bare soil, and the spatial resolution is 5 m × 20 m [11]. Satellite remote sensing data is limited by low spatial resolution and low temporal resolution, and it is difficult to improve the real-time and accuracy of monitoring bare soil moisture content. At the same time, the thermal inertia method is mainly used in bare soil or low vegetation, which is limited to the monitoring of soil surface moisture. In the semi-arid region with sparse vegetation, MODIS image is used to monitor soil moisture, which is easily affected by the actual situation [12]. The GPS (Global Positioning System) satellite signal-to-noise ratio soil moisture inversion result is good, and the experimental data processing results are consistent with the theory. However, the detection process is affected by surface roughness and vegetation cover, and there are still great difficulties [13]. Satellite remote sensing data are affected by vegetation cover, and the effectiveness of data cannot be well guaranteed in the inversion model of bare soil moisture.

With the development of airborne remote sensing and the diversification of UAV sensor types, UAV remote sensing has become a new remote sensing application platform. It has the characteristics of ultra-high resolution, high frequency acquisition and real-time observation in the refinement of regional information, which can complement the satellite remote sensing ability, alleviate the contradiction between high spatial resolution and high temporal resolution, and achieve the unification of space and time on the basis of low cost [14].

The convenience, flexibility and practicality of UAV remote sensing have played a good role in promoting the development of bare soil moisture content monitoring. Using the UAV remote sensing platform to carry the thermal infrared imager to quickly obtain the regional surface temperature, and then determine the regional water regime has gradually become a hot spot, providing a new idea for monitoring soil moisture content. Many scholars in China and abroad used UAV thermal infrared technology to monitor crops, quickly and efficiently estimated crop water stress index (CWSI), and found the correlation between CWSI and soil moisture [15], [16], [17]. With the development of science, Egea et al. used UAV thermal infrared technology to collect the image information of olive orchard, and realized the monitoring of soil moisture in olive orchard through image processing [18]. Sun Sheng et al. established the soil moisture prediction model of walnut orchard by thermal infrared image, and realized the transformation from theoretical model to practical application, from single plant level to regional scale [19]. With the help of UAV thermal infrared remote sensing platform, the inversion of soil moisture content of vegetation cover is closely related to CWSI of crops and canopy temperature of trees.

In terms of bare soil moisture content monitoring, the UAV multi-spectral remote sensing technology was used to establish the regression models for the inversion of soil moisture content with different spectral reflectance factors in different bands by using the partial least squares regression method, stepwise regression method and ridge regression method for bare soil moisture content with clay soil texture, and the quantitative relationship was analyzed to determine the optimal monitoring depth [20]. The experimental soil samples of the dump in the mining area were selected to form the experimental soil column samples, which were injected regularly. The soil spectral reflectance and soil water content in each band were analyzed, and it was concluded that the multi-spectral G-R-N-BP model had high inversion accuracy and strong availability, which provided good support for surface water monitoring of bare soil in the mining area [21]. The L-band active and passive microwave was used, supplemented by surface roughness information to invert the bare surface soil moisture [22]. UAV thermal infrared remote sensing monitoring of bare soil moisture content has not been fully developed. The method of extracting temperature information from the image is simple and mature because of the single ground object in the thermal infrared image of bare soil. So the application of UAV thermal infrared remote sensing in large-scale monitoring of bare soil moisture content is feasible.

In this paper, the surface bare soil is taken as the research object. A UAV visible light camera and thermal infrared instrument are used to obtain the bare soil area of a visible light image and thermal infrared image. The relationship between radiation value and temperature is analyzed by quadratic and quartic polynomial regression, and the telemetry surface temperature (brightness temperature) is retrieved. The relationship model between measured temperature and measured soil moisture content was analyzed by linear regression, which preliminarily confirmed the feasibility of inversion of bare soil moisture content by temperature. Then, the linear model is used to fit the regression analysis of large-area brightness temperature and moisture content to determine the inversion models in different periods, and the quantitative accuracy is verified, so as to infer the optimal inversion period and inversion model.

Section snippets

Study area overview

The experiment area is located in the Huyu Campus of Taiyuan University of Technology, Wanbailin District, Taiyuan City, Shanxi Province (37°51′ N, 112°31′ E, 786 m above sea level). The area has a typical temperate continental monsoon climate with four distinct seasons. The diurnal temperature difference is large. The rainy season lasts three months (July–September) and the average annual precipitation is approximately 480–500 mm. Sunshine is plentiful throughout the year, a total of 2800 h.

Correlation analysis between gray value of thermal infrared radiation and measured temperature

Using the theoretical analysis described, we can calculate the thermal infrared radiation emitted by the target in an ideal state based on the temperature value; the method can also inverse the target temperature if the radiation value is known. The relationship between the target thermal infrared radiation value obtained by the equipment and the measured soil temperature was also proved.

Fig. 5 is a linear fitting model between the gray value of thermal infrared radiation and the measured

Discussion

All substances emit infrared radiation if their temperature exceeds absolute zero. The thermal infrared radiation corresponding to 273–333 K in the study area was obtained by a D-TIRC1000 thermal infrared instrument. The blackbody radiation value at the corresponding temperature was obtained by multiplying the integral operation value of Planck’s blackbody radiation law and the spectral band response function. A quadratic polynomial model is used to regress gray value of thermal infrared

Conclusion

A conversion formula for gray value of thermal infrared radiation and temperature was obtained by calculating Planck’s blackbody radiation law and the band response function of the spectrum. The gray value of thermal infrared radiation exhibits a good linear correlation with the measured temperature. Therefore, the gray value of thermal infrared radiation can be used to predict the temperature (i.e., brightness temperature). It is feasible to use UAV thermal infrared remote sensing technology

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

The study was supported by the Natural Science Foundation of Shanxi Province, China (Grant No. 201901D111074), and National Natural Science Foundation of China (Award Number: 42101414 and 51704205).

References (25)

  • Q.K. Cai et al.

    Soil Moisture estimation of wheat field based on theoretical dry-wet edge and improved TVDI

    Trans. Chin. Soc. Agric. Mach.

    (2020)
  • L. Zeng et al.

    A three-variables cokriging method to estimate bare-surface soil moisture using multi-temporal, VV-polarization synthetic-aperture radar data

    Hydrogeol. J.

    (2020)
  • Cited by (0)

    View full text