A Deep-Learning Approach to Soil Moisture Estimation with GNSS-R
نویسندگان
چکیده
GNSS reflection measurements in the form of delay-Doppler maps (DDM) can be used to complement soil from SMAP Mission, which has a revisit rate too slow for some hydrological/meteorological studies. The standard approach, only considers peak value DDM, is subject significant amount uncertainty due fact that DDM not affected by moisture, but also complex topography, inundation, and overlying vegetation. We hypothesize information entire 2D could help decrease under various conditions. application deep-learning-based techniques potential extract additional while simultaneously allowing incorporation contextual external datasets. This work explored data-driven approach convolutional neural networks (CNNs) determine relationships between measurement surface parameters, providing groundwork mechanism achieve improved global moisture estimates. A CNN was trained on CYGNSS DDMs ancillary datasets as inputs, with aligned values targets. Data were aggregated into training sets, developed process them. Predictions studied using an unbiased subset samples, showing strong correlation target values. With this network, product generated 2017–2019 generally comparable existing products, shows advantages spatial resolution coverage over regions where does perform well. Comparisons in-situ demonstrate network predictions ground truth high temporal resolution.
منابع مشابه
Soil Moisture Monitoring Using GNSS-R Signals; First Experimental Results with the SAM Sensor
Observing the Earth surface with Global Navigation Satellite Systems (GNSS) reflected signals has become a noteworthy remote sensing technique for the scientific community. The growing interest in GNSS as a remote sensing tool is due to its global availability and the carrier frequencies used. In fact, L-band, in which all current and next-future Global Navigation Satellite Systems emit, is a p...
متن کاملUse of GNSS SNR data to retrieve soil moisture and vegetation variables over a wheat crop
This work aims to estimate soil moisture and vegetation characteristics from Global Navigation Satellite System (GNSS) Signal to Noise Ratio (SNR) data using direct and reflected signals by the land surface surrounding a ground-based antenna. Observations are collected over a rainfed wheat field in southwestern France. The retrievals are compared with two independent reference datasets: in situ...
متن کاملMelanoma detection with a deep learning model
Background: Skin cancer is one of the most common forms of cancer in the world and melanoma is the deadliest type of skin cancer. Both melanoma and melanocytic nevi begin in melanocytes (cells that produce melanin). However, melanocytic nevi are benign whereas melanoma is malignant. This work proposes a deep learning model for classification of these two lesions. Methods: In this analytic s...
متن کاملSoil Moisture Estimation Using Remote Sensing
Knowledge of soil moisture content in the root zone is important throughout a wide range of environmental applications, yet adequate monitoring or modelling of this parameter, particularly at larger spatial scales, is difficult due to its high spatial and temporal variability. To overcome the land surface model limits on soil moisture estimation accuracy, point measurement spatial coverage limi...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
ژورنال
عنوان ژورنال: Remote Sensing
سال: 2022
ISSN: ['2315-4632', '2315-4675']
DOI: https://doi.org/10.3390/rs14143299