Сибирский федеральный университет

Тип публикации: статья из журнала

Год издания: 2022

Идентификатор DOI: 10.1016/j.scitotenv.2021.151121

Ключевые слова: sentinel-1, uav, digital elevation model, radar backscattering, artificial neural network, soil moisture

Аннотация: Soil surface moisture is one of the key parameters for describing the hydrological state and assessing the potential availability of water for irrigated plants. Because the radar backscattering coefficient is sensitive to soil moisture, the application of Sentinel-1 data may support soil surface moisture mapping at high spatial resolution by detecting spatial and temporal changes at the field scale for precision irrigation management. This mapping is required to control soil water erosion and preferential water flow to improve irrigation water efficiency and minimise negative impacts on surface and ground water bodies. Direct observations of soil surface moisture (5-cm thickness) were performed at an experimental plot in the study site of the All-Russian Scientific Research Institute of Irrigated Agriculture, near the village Vodnyy, Volgograd region. Soil surface moisture retrieval from Sentinel-1 was performed at the same location. A second set of soil surface moisture was calculated for the soil sampling sites using the permittivity model, based on the estimates of soil surface characteristics: a) reflectivity, obtained by the neural network method from Sentinel-1 observations; b) roughness, obtained from the geodata of the stereoscopic survey with unmanned aerial vehicle Phantom 4 Pro. The raster set of soil surface moisture geodata was obtained based on the reflectivity geodata raster set to solve the inverse problem using a permittivity model that considers the soil texture of the experimental plot. The determination coefficient (0.948) and standard deviation (2.04%) were obtained by comparing both sets of soil moisture point geodata taken from the same soil sampling sites. The values confirmed a satisfactory linear correlation between the directly measured and indirectly modelled sets. A comparison of the two sets of geodata indicated a satisfactory reproduction of the first set by the second set. As a result, the developed method can be considered as the scientific and methodological basis of the new tech-nology of soil surface moisture monitoring by radar, which is one of the basic characteristics used in precision ir-rigation management. (c) 2021 Elsevier B.V. All rights reserved.

Журнал: SCIENCE OF THE TOTAL ENVIRONMENT

Выпуск журнала: Vol. 807, Part 2

Номера страниц: 151121

Место издания: ELSEVIER

Издатель: 3

• Zeyliger A.M. (Russian State Agr Univ, Moscow Timiryazev Acad, Moscow, Russia)
• Muzalevskiy K.V (Siberian Fed Univ, RAS, Siberian Branch, Krasnoyarsk Sci Ctr,Div Fed Res Ctr,Kirensky Inst, Krasnoyarsk, Russia)
• Zinchenko E.V (All Russian Sci Res Inst Irrigated Agr, Volgograd, Russia)
• Ermolaeva O.S. (Russian State Agr Univ, Moscow Timiryazev Acad, Moscow, Russia)

• Web of Science Core Collection