Understanding the vertical distribution of soil moisture is crucial for making informed decisions in various applications, ranging from precision agriculture to hydrological modeling. Four machine learning algorithms, including random forest, extreme gradient boosting, deep learning, and support vector regression were employed to estimate the soil moisture profile from collected tower-based L-band and P-band brightness temperature observations in Victoria, Australia. The results showed that random forest outperformed the other algorithms, with root mean square error (RMSE) values of 0.03, 0.04, and 0.06 m3/m3for depths of 0-5 cm, 0-30 cm, and 0-60 cm, respectively

Soil moisture is a key state variable in environmental monitoring and in the water, energy and carbon cycles [1] - [3] . Farm management decisions, including the timing of planting, application of fertilizers, pesticides, herbicides, and irrigation scheduling, are influenced by soil moisture status [4] , [5] . Moreover, it is highly variable both in space and time and the estimation of this variable is challenging because the amount of moisture in the surficial soil layer is influenced by soil texture [6] , [7] . Passive microwave remote sensing is a well-accepted technique for estimating soil moisture due to the large contrast between the dielectric properties of liquid water (~80) and that of dry soil matter (~3.5) [8] , and reduced sensitivity to surface roughness and vegetation as compared to active microwave [9] . Current missions, including the Soil Moisture and Ocean Salinity (SMOS; [10] ) and Soil Moisture Active Passive (SMAP; [11] ) operating at L-band radiometer (~21 cm) are only able to detect shallow soil moisture (up to 5 cm in depth; [12] ). Compared with L-band, P-band (~ 40 cm) is even less sensitive to vegetation water content [13] and surface roughness [14] , and is able to penetrate deeper into the soil providing information about moisture over deeper depths (~10 cm; [12] ).

This study aims to discuss the use of L-band radar backscatter and phase information to analyze soil moisture (SM) and crop conditions, focusing on corn vegetation. While previous research has primarily utilized total backscatter magnitude to assess SM and crop conditions, this work explores the potential of phase information, which may be more sensitive to crop structure and suitable for monitoring crop dynamics. The methodology involves radar measurements using the University of Florida L-band Automated Radar System (UF-LARS), destructive vegetation sampling, and calibration techniques. Preliminary results demonstrate significant differences in CPD between bare soil and vegetated conditions, with phase variations associated with different growth stages of corn. This ongoing project indicates the potential of phase information to characterize vegetation growth stages for corn and similar crops.

Timely, frequent, and complete cropland soil moisture information acquired throughout the growing season is critical for agricultural policy, production, food security, and food prices. The NASA Soil Moisture Active and Passive (SMAP) mission provides a reliable data source for cropland soil moisture assessment. This paper presents Crop-CASMA - a web GIS application tool for cropland soil moisture monitoring and assessment based on SMAP data. This interactive Web service-based GIS application tool enables CONUS SMAP derived soil moisture data visualization, dissemination, and analytics. In this paper, we describe the Crop-CASMA application system architecture, the application implementation, and the data it serves. In addition, we also present a few snapshots of the Crop-CASMA data for cropland soil moisture monitoring. The release of Crop-CASMA greatly enhances the user experience and facilitates using soil moisture data products for crop condition monitoring and decision support.

The latest developments in radar mission concepts suggest that subdaily synthetic aperture radar will become available in the next decades. The goal of this study was to demonstrate the potential value of subdaily spaceborne radar for monitoring vegetation water dynamics, which is essential to understand the role of vegetation in the climate system. In particular, we aimed to quantify fluctuations of internal and surface canopy water (SCW) and understand their effect on subdaily patterns of L-band backscatter. An intensive field campaign was conducted in north-central Florida, USA, in 2018. A truck-mounted polarimetric L-band scatterometer was used to scan a sweet corn field multiple times per day, from sowing to harvest. SCW (dew, interception), soil moisture, and plant and soil hydraulics were monitored every 15 min. In addition, regular destructive sampling was conducted to measure seasonal and diurnal variations of internal vegetation water content. The results showed that backscatter was sensitive to both transient rainfall interception events, and slower daily cycles of internal canopy water and dew. On late-season days without rainfall, maximum diurnal backscatter variations of >2 dB due to internal and SCW were observed in all polarizations. These results demonstrate a potentially valuable application for the next generation of spaceborne radar missions.

Accurate knowledge of soil and crop conditions such as water content, biomass, and phenology is crucial in agriculture for estimating growth and productivity. Remote sensing observations at microwave frequencies are sensitive to various soil and crop characteristics. For example, both the active (radar) and passive (radiometer) microwave sensors measure radiation quantities that are functions of soil and vegetation dielectric constant and exhibit sensitivities to soil moisture (SM) and vegetation water content (VWC). In addition to SM, the quantities are also influenced by other land surface parameters such temperature, soil surface roughness, vegetation geometry. Brightness temperature (TB) is more sensitive to SM and less sensitive to surface roughness and vegetation geometry compared to radar backscattering ( σ0). Satellite passive observations have been widely used because of their high temporal resolutions (frequency of every 3 days), but these observations are available at coarse resolutions of about 25-40km. In contrast, satellite active observations from synthetic aperture radar (SAR) provide finer resolution (lower than 3 km), but low temporal resolution. However, with the recent availabilities of both Radarsat-2 and Sentinel-1, we have almost weekly observations combining both observations. In addition, upcoming launch of NISAR mission in 2021 will provide unprecedented opportunities to harness multifrequency observations at L-, S-, and C-bands. In this study, we investigate sensitivities of such multifrequency observations to soil and vegetation water content using current C-band SAR data and ground-based L-band data (in preparation for NISAR) obtained from field experiments. Active L-band observations are the most sensitive to SM variations even when the biomass in agriculture fields such as corn is high; in contrast, active C-band observations are more sensitive to vegetation.

The SMAP (Soil Moisture Active/Passive) satellite provides global soil moisture (SM) estimates that can be used for scientific research and applications (such as the hydrological cycle, agriculture, ecology, and land atmosphere interactions). Currently, SMAP provides the enhanced radiometer-only SM product (L2SMP) at 9 km grid resolution. However, this spatial resolution is still not enough to satisfy the needs of some studies that require a finer spatial resolution SM product, particularly in agricultural and watershed applications. This study applied a downscaling algorithm to the SMAP 9 km SM product to produce a 1 km resolution over the CONUS (Contiguous United States). The downscaling algorithm is based on the relationship between temperature change and SM modulated by Normalized Difference Vegetation Index (NDVI) of a given time period. This relationship was modeled using variables derived from NLDAS (North America Land Data Assimilation System) and NASA's LTDR (Land Long Term Data Record) between 1981 - 2018. The algorithm was implemented uses the 1 km MODIS Aqua LST (Land Surface Temperature) product. The downscaled SMAP 1 km SM was validated using in situ SM measurements from the ISMN (International Soil Moisture Network). The validation metrics show an improved overall accuracy of the downscaled SM.

In this paper, we developed and evaluated a phenology-based coherent scattering model to estimate terrain backscatter at the L-band for growing corn. The scattering model accounted for combined effects from periodicity in soil and vegetation, and changes in plant structure and phenology. The model estimates were compared with observations during the two growing seasons in North Central Florida. The unbiased average root-mean-square (rms) differences between the model and observations decreased from 5 to 1.31 dB when these combined effects were included. During the early stage, direct scattering from soil was the primary scattering mechanism, and as the vegetation increased, the interactions between stems and soil became the dominant scattering mechanism. The most sensitive soil parameters were moisture content and rms height, and vegetation parameters were the widths of stems, leaves, and ears, and the stem water content. This paper demonstrates that it is necessary to consider periodicity and plant structural effects in algorithms to retrieve realistic soil moisture in agricultural terrain.

In this study, a data-fusion algorithm is developed for estimation of high-resolution brightness temperatures (TB) at 1km from Soil Moisture Active Passive (SMAP) fine-grid TBproduct at 9km. It uses image segmentation to spatio-temporally cluster the study region based on meteorological and land cover similarity, followed by a support vector machine based regression that computes the value of the high-resolution TBat all pixels. High resolution remote sensing products such as land surface temperature, normalized difference vegetation index, enhanced vegetation index, precipitation, soil texture, and land-cover were used for disaggregation. The algorithm was implemented in Iowa, United States, from May to September 2016, and compared with the field observations of TBfrom Microwave Water and Energy Balance Experiment conducted as a part of the Soil Moisture Active Passive Validation Experiment (SMAPVEX16-MicroWEX). Additionally, they were also compared with the Sentinel downscaled SMAP TBat 1km. High resolution soil moisture is subsequently derived from high resolution TBusing inverse models.

In this study, the impact of spatial variability due to the heterogeneity of vegetation in the agricultural region on passive microwave signatures available at various scales are explored using the brightness temperature (TB) observed from ground, air, and space. These observations were conducted during a growing season of corn and soybean in South Fork watershed, Iowa, as part of the NASA-Soil Moisture Active Passive Validation Experiment (SMAPVEX16). Both empirical and physically-based microwave emission models are used to understand the effects of vegetation on TBfor corn and soybean using ground-based TBobservations. The modeled TBwill be upscaled based upon the USDA crop layer map to compare with the TBobserved in the coarse scales.

Soil moisture (SM) is an important land surface variable for understanding the water cycle, ecosystem productivity, and linkages between water-carbon cycles. For SM studies, observations at L-band frequencies are more desirable due to larger penetration depths. The NASA Soil Moisture Active/Passive (SMAP) mission includes active and passive sensors at L-band to provide global observations of SM. The active observations are available from April-July 2015. In addition to the SM sensitivity, radar backscatter is highly sensitive to roughness of soil surface and scattering within the vegetation. Despite much progress in the development of backscattering models, there is still a gap in validating such models under dynamic vegetation conditions such agricultural crops. The goal of this study is to validate an incoherent model using season-long active observations over a soybean field at high temporal resolution during the SMAPVEX16-MicroWEX experiment.

Soil moisture (SM) is an important land surface variable for understanding the water cycle, ecosystem productivity, and linkages between water-carbon cycles. For SM studies, observations at L-band frequencies are more desirable due to larger penetration depths. The NASA Soil Moisture Active/Passive (SMAP) mission includes active and passive sensors at L-band to provide global observations of SM. The active observations are available from April-July 2015. In addition to the SM sensitivity, radar backscatter is highly sensitive to roughness of soil surface and scattering within the vegetation. Despite much progress in the development of backscattering models, there is still a gap in validating such models under dynamic vegetation conditions such agricultural crops. The goal of this study is to improve a coherent model and evaluate it using season-long active observations at high temporal resolution during the SMAPVEX16-MicroWEX experiment.

Global Navigation Satellite Systems-Reflectometry (GNSS-R) is an emerging remote sensing technique that makes use of navigation signals as signals of opportunity in a multi-static radar configuration, with as many transmitters as navigation satellites are in view. GNSS-R sensitivity to soil moisture has already been proven from a ground-based and airborne experiments, but studies using space-borne data are still preliminary. This work presents a sensitivity study of Using TechDemoSat-1 GNSS-R data to soil moisture over different types of surfaces (i.e. vegetation covers). Despite the scattering in the data, which can be attributed to the temporal and spatial (footprint size) collocation mismatch with the SMOS and MODIS NDVI data, and errors in the land use data preliminary results show a good correlation with soil moisture.

In this paper, bias correction approaches are investigated to understand their impact on assimilating active and/or passive microwave observations on near-surface soil moisture (SM) estimates. Synthetic and field observations were assimilated in a soil-vegetation-atmosphere transfer model linked with an integrated active-passive model at L-band for bare soil. The two bias correction methods included in this study are the online bias correction with feedback (BCWF) with extended implementation with nonlinear observation operators and the simultaneous state parameter (SSP) update. New equations for BCWF were derived for the case of nonlinear observation operators because current versions of this approach were not applicable for improving SM by assimilating microwave observations. In SSP, the bias is compensated by tunning the values of the parameters. The two approaches resulted in similar accuracy for improving SM estimates compared with the uncorrected estimates. SSP showed the highest certainty for both synthetic and field observations. Using the bias correction methods, the mean estimates of SM improved by up to 88%, 87%, and 94%, when passive, active, and active-passive synthetic observations were assimilated, respectively, compared with the open-loop estimates. In contrast, when assimilating field observations from the Eleventh Microwave Water Energy Balance Experiment, the mean estimates of SM improved by up to 44%, 18%, and 48%, when passive, active, and active-passive observations were assimilated, respectively, compared with open-loop estimates. The decrement in improving the SM estimates suggests sources of uncertainty other than those from model parameters and forcings.

The ground-based University of Florida L-band Automated Radar System (UF-LARS) was developed to obtain observations of normalized radar backscatter (\mmbσ0) at high temporal resolution for soil moisture applications. The system was mounted on a 25 m manlift with capabilities of antenna positioning for multi-angle data acquisition and ranging. The RF subsystem of UF-LARS was based upon the established designs for ground-based scatterometers employing a vector network analyzer with simultaneous acquisition of V- and H-polarized returns. System integration and automated data acquisition were enabled using a software control system. Fifteen-minute observations of \mmb σ0collected over a growing season of sweet-corn and bare soil conditions in North Central Florida, were used to study the sensitivity of \mmbσ0to growing vegetation and near-surface (0-5 cm) soil moisture (\mmbSM0 - 5). On average, \mmb σ\mmbVV0were observed to be 23% higher than \mmbσ\mmbHH0during the mid- and late-stages of crop growth due to the vertical structure of stems. The correlation between 3-day observations of \mmbSM0 - 5 and \mmbσ\mmbVV0reduced by 55% compared to those obtained for ≤ 30-min observations. These findings suggested that data set at high temporal frequencies can be used to develop more realistic and robust forward backscattering models.

In this study, sensitivity of active and passive (AP) observations at L-band to near-surface SM was analyzed for bare sandy soils. The complementarity of AP microwave observations was used to obtain realistic SM profile that matched well with both AP observations during dynamic moisture conditions. Active observations exhibit less sensitivity to SM changes and higher sensitivity to surface roughness than passive observations. Based upon these findings, the observed brightness temperatures (TBs) were used to estimate a SM profile using an emission model. The backscatter (σ°) observations were used to estimate root mean square height (s) and correlation length (cl) using a backscatter model. The estimated SM profile, s, and cl resulted in RMSDs of 4.55K and 0.81dB between the observed and modeled TBand σ° values, respectively, for the rough surface. This study demonstrates the integrated use of AP to improve SM estimates.

The L-band signals broadcast by GPS satellites are attenuated by vegetation, making it problematic, if not impossible, to predict the performance of the system in forested areas without some quantitative measure of the structure and density of the local forest canopy. Airborne laser swath mapping (ALSM) observations can be used to rapidly and remotely sample the structure and density of forested areas. We report here the results of a study performed to determine the attenuation of GPS signals in forests, by correlating changes in the signal-to-noise ratio (SNR) of the received GPS signals under different canopies, using three dimensional structure and density information about each canopy derived from ALSM observations. The results of this study verify that the loss of signal is strongly correlated with the local structure and density of the forest, and we demonstrate how the ALSM point cloud can be used to better predict the attenuation of the GPS signals. The results of this research also pertain to other modes of microwave transmission in forested areas, including satellite and cellular telephony, and the estimation of biomass from L-band radar.