SMAP’s polarization index (PI) is correlated to the mass of water in vegetation above the soil surface. Vegetation emits an unpolarized signal, while the soil emission is stronger in vertical polarization than horizontal. Thus, a more polarized SMAP measurement is indicative of less water in the vegetation canopy. We hypothesize that less vegetation canopy water in the afternoon is indicative of crops that are not water-stressed, as water in vegetation is transpired in the morning and then recharges at night. We tested the diurnal variation in SMAP polarization index in the Corn Belt state of Iowa in 2020 and 2021 at two locations: one under severe drought and an area not under drought conditions. The mean diurnal change in PI was significantly positive at the site not under drought conditions (p = 0.0250), while the diurnal cycle at the site experiencing drought was not significant (p = 0.1587). These tests were repeated for the eight surrounding SMAP pixels at both sites, and similar results were found. Filtering out rainy periods resulted in a sample size that was too small to support significant results.

Evaluation of soil moisture information should consider when key crop development stages occur and, ultimately, when farmers must make decisions based on soil moisture status. Therefore, we assessed the performance of three microwave satellites (SMAP, SMOS, and METOP/ASCAT) and three reanalysis models (MERRA-2, NARR, and WEPP) in the U.S. Corn Belt in the context of agricultural management. Thermal time and crop progress reports from the USDA-NASS defined critical transition periods of crop growth and management decisions for the validation process. Contrary to calendar timelines like annual segments, these key events separate the year into five irregular segments. Measurements at the South Fork Core Validation Site from 2016 to 2020 show that the two passive microwave satellites are dry compared to in-situ observations, but the active satellite and reanalysis model products were almost always wetter. Overall, most products perform the best during the pre-planting and post-harvest segments when no crops are present and worst during the active management phase when crops are starting to grow.

Attenuation of microwave radiation by a vegetation canopy can be quantified by the vegetation optical depth (VOD). VOD is directly proportional to the total mass of liquid water contained within vegetation tissue per ground area. Traditionally this quantity has been called the vegetation water content (VWC). However, it is an extensive property and other terms characterizing the liquid water in both soil and vegetation that use the word "content" are all intensive properties. I propose that the microwave remote sensing community instead use the term vegetation canopy water or VCW because the same three letters in the acronym are retained.

In croplands, the L-band terrestrial brightness temperature is a function of not only soil moisture and vegetation, but also time-varying soil surface roughness. Soil surface roughness changes in response to human activities, such as the planting of crops, soil tillage, and rainfall. We use in situ data from the South Fork SMAP Core Validation Site in the US Corn Belt to determine the magnitude and polarization dependence of the soil surface roughness signal at the satellite scale. We find that when crops are not present, soil surface roughness retrievals are larger than anticipated, and are effectively independent of polarization except for their largest values.

Satellite L-band vegetation optical depth (L-VOD) contains new information about terrestrial ecosystems. However, it has not been evaluated against the geophysical variable that it represents, plant water, the mass of liquid water contained within vegetation tissue per ground area. We quantitatively assess the seasonal variation of three L-VOD products at the South Fork Core Validation Site in the Corn Belt state of Iowa where L-VOD is directly proportional to crop plant water. We use three satellite-scale crop plant water estimates:in situmeasurements; a normalized difference water index (NDWI) calibrated within situmeasurements; and a crop model. We find that overall the L-VOD satellite products are 0.02–0.09 Np (0.4–${1.7} \,\,\text {kg} \cdot \text {m}^{-2}$) lower than the three estimates. We show that overestimation of L-VOD can be attributed to dynamic soil surface roughness, and hypothesize that crop plant water observations will require the incorporation of this effect into retrieval algorithms.

In the summers of 2018 and 2019, a 433 MHz radio link was installed in an Iowa corn field, intended to measure the vegetation's water content. A set of three transmitters within the canopy emitted microwave radiation omnidirectionally in succession to each other, while the receiver, also within the corn canopy, listened continuously. The system was in operation for 68 days in 2018, beginning when the corn was only 0.3 m tall and ending in a full canopy environment. An increasing vegetation water content corresponded to an increasing canopy dielectric constant and thus a lower canopy transmissivity. A weakened signal received was for the most part attributed to electric field attenuation. In 2019, antennas remained in the field through October, observing the dry-down senescence as an increasing signal strength. The magnitude of air - soil reflection is another strong factor on received signal strength, as for all transmitter - receiver pairs the direct and ground - reflected signal destructively interfered. Radiative transfer models of the system often simulate the received signal to be stronger than observed. Neglecting scattering may be the cause of this error. Utilizing the software Signals of Opportunity Coherent Bistatic Scattering Simulator (SCoBi) may produce a more accurate calculation of power received.

Satellite L-band vegetation optical depth (L- VOD) is a relatively new but potentially valuable vegetation product. We hypothesize that the relationship between L- VOD and crop plant water, as characterized by the “b-parameter,” should change over the growing season as crops progress through different developmental stages. We find that the b-parameter derived from SMOS and SMAP L- VOD and satellite-scale estimates of crop plant water made by the Agro-IBIS model does in fact exhibit a growing-season pattern. This pattern is consistent among different years during the first half of the growing season.

A 433 MHz radio link was installed in a central Iowa corn canopy just after the crop's emergence to measure the amount of water stored in vegetation. Three transmitters, which the corn would grow above, where separated from the receiver by a horizontal distance of 50 m. Antennas operated for 68 days during the summer of 2018, while plants grew from a height of 0.3 m to 2.7 m. Water content of the corn by an observed weakened signal incident on the receiver. An existing empirical model for corn dielectric constant was used when modeling signal strength, applying in situ vegetation measurements. Interference with the ground-reflected signal was accounted for; the ground signal acts destructively with the direct propagation path in this experiment. Early in the growing season, signal strength is modeled rather well. When ears were present, the modeled signal is much stronger than the observed. This may be the result of neglecting scattering, which may be significant as the radiation was vertically polarized and likely strongly affected by the large vertical stems of the crop.

NASA's Soil Moisture Active Passive (SMAP) is a satellite L-band radiometer whose primary mission is to measure soil moisture. However, it also retrieves vegetation optical depth (VOD), the degree to which vegetation attenuates microwave radiation. We believe that VOD has the potential to monitor crop growth in the US Corn Belt. To show its value, we compare SMAP VOD to satellite-scale estimates of crop productivity created using the Agricultural Integrated BIosphere Simulator (Agro- IBIS) and observed weather at the South Fork SMAP Core Validation Site in the Corn Belt state of Iowa. We find that SMAP VOD is directly proportional to crop water, the mass of liquid water in crop tissue. We discovered this relationship by using new empirical models that relate crop water to crop dry mass, the standard output for Agro-IBIS as well as all other crop models, created with in situ data spanning multiple years and stages of crop development. The value of the proportionality constant (or "b-parameter") relating VOD to crop water at the satellite scale is half as large as the SMAP value. Because L-band VOD is directly proportional to crop water at the satellite scale, SMAP has the potential to evaluate the large-scale performance of crop models in the Corn Belt on a near daily basis.

SMAP soil moisture retrieval currently relies on a relatively smooth parameterization of soil surface roughness in croplands. However, in agricultural regions like the U. S. Corn Belt where tillage is common, roughness varies according to farm management practices, increasing due to harvest and tillage and decreasing from field cultivation and rainfall. We approximate roughness at the South Fork core validation site during 2016 by re-arranging the SMAP Single Channel Algorithm to retrieve HR from observed brightness temperature when in situ observations of soil moisture and ancillary data are provided. The result is a temporally dynamic HR, rougher than the SMAP default parameterization, with a slight sensitivity to soil moisture. This analysis is believed to be the first to measure a temporally dynamic HR for croplands at the satellite-scale. We hypothesize that SMAP performance will improve in the U. S. Corn Belt when the refined HR is utilized during soil moisture retrieval.

The European Space Agency's Soil Moisture and Ocean Salinity (SMOS) satellite has recently been shown to measure variables containing information relevant to agronomists. SMOS was initially intended to monitor the water content of soil. However, a combination of SMOS's antenna technology and data processing algorithms make it possible to estimate the mass of water contained in vegetation tissue. Recent work by Hornbuckle et al., as well as Lawrence et al., suggest τ roughly mirrors the growth and senescence of crops [1, 2]. In this paper we analyze SMOS data from an intensively cultivated agricultural region in the Midwest to provide new information about crop phenology. In addition to modeling the seasonal pattern of crop growth, we estimate the day of the year when τ reaches its peak. Because SMOS has a fine temporal resolution, the ability to model τ during a growing season could be useful to understanding changes in crop development, climate conditions, as well as forecasting future growth cycles.

It is believed that the cosmic-ray neutron sensor can be used to measure vegetation. We compare biomass water equivalent (BWE) to neutron counts from a cosmic-ray neutron sensor. We believe that the thermal neutrons will have a highly correlated inverse relationship with BWE. This is due to the ambient energy of thermal neutrons and the ability of hydrogen to capture them easily. Our results for soybean in 2017 showed that there was better correlation between raw fast neutron count (R2= 0.61) and BWE than raw thermal neutron count (R2= 0.21). It would be ideal to explore this relationship of fast neutrons to BWE, with the addition of other hydrogen pools at the surface.

The Soil Moisture Ocean Salinity (SMOS) and Soil Moisture Active Passive (SMAP) Level 2 Soil Moisture products both exhibit a dry bias over agricultural regions. In regions such as the U.S. Corn Belt, where vegetation water content is high during the growing season and near zero in the winter' the year can be split into periods where retrieved optical thickness is either representative of vegetation water content or surface roughness. We hypothesize that allowing roughness to vary with retrieved optical thickness outside of the growing season will improve the dry bias in the U.S. Corn Belt. Pixels that have a distinct boundary between rough soil and vegetated conditions need to be identified to determine where this modified retrieval process could be useful. SMOS auxiliary land surface fractions are used as a filter for forest, urban, and open water before visually inspecting timeseries of optical thickness for roughness-vegetation patterns.

The European Space Agency's SMOS and NASA's SMAP soil moisture remote sensing satellite missions do not perform well in several agricultural regions. For example, we have found that SMOS and SMAP soil moisture retrievals are 0.09 to 0.07 m3m-3lower, respectively, than an in-situ soil moisture network in one such agricultural region in Central Iowa, USA. We hypothesize that this dry bias is in part caused by an incorrect parameterization of the effect of soil surface roughness, the mm-scale variations in the height of the soil surface. A large field campaign called SMAPVEX16-IA was held in this region in 2016. As part of the campaign, we measured soil surface roughness at nearly 20 sites within the domain of a SMOS/SMAP satellite pixel. We found that our observed soil surface roughness was higher than what is used by the SMOS soil moisture retrieval algorithm. If the SMOS and SMAP retrieval algorithms were adjusted to account for a higher soil surface roughness, this would increase retrieved soil moisture and decrease the soil moisture bias. We also found that our measurements of soil surface roughness are consistent with changes in the τ parameter, which is sensitive to soil surface roughness, retrieved by both the SMOS and SMAP missions.

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.

The Soil Moisture Ocean Salinity (SMOS) soil moisture retrievals are too dry and noisy when compared to the South Fork of the Iowa River (SFIR), a heavily agricultural watershed with a USDA-ARS in situ soil moisture network. After testing for invalid retrievals, errors in auxiliary datasets, and a non-representative parameterization of scattering in the canopy, soil surface roughness changes were identified as the next potential source of the dry bias. Soil surface roughness increases the amount of radiation emitted from the surface; if the SMOS processor does not “know” that the soil is rough, the increased brightness temperature is interpreted as a drier soil. When the processor does account for a rougher soil surface, the previous dry bias between SMOS and the SFIR will reduce. This ability to zero the bias comes at a high cost: SMOS sensitivity to the SFIR soil moisture is halved between the moderately-rough and rough scenarios.

The Soil Moisture Active Passive (SMAP) mission may require ancillary vegetation optical thickness (τ) information as part of SMAP's soil moisture retrieval algorithm. Currently, the ancillary τ data comes from a normalized difference vegetation index (NDVI) climatology that is converted to τ. The Soil Moisture Ocean Salinity (SMOS) satellite measures τ as part of its soil moisture retrieval algorithm. In this paper, we compare SMOS τ to SMAP's proposed NDVI climatology-derived τ (SMAP τ). During the growing season in heavily cultivated parts of Iowa, SMAP τ is usually larger than SMOS τ. The timing of the peak in τ is similar between the two data sets on the whole, but some SMOS pixels in 2012 peak earlier by 15-20 days.

The European Space Agency's Soil Moisture and Ocean Salinity (SMOS) satellite mission provides microwave L-band measurements of vegetation optical thickness over the Earth. Optical thickness is related to water held in vegetation. The water content of crops varies over the growing season from a minimum during planting to a maximum during reproduction and back to a minimum during senescence. We found that in Iowa in 2010 the change in SMOS optical thickness over the growing season can be related to crop yields. However, there are inconsistencies in the optical thickness data, particularly high-frequency variation and unexpected changes outside of the growing season. We hypothesize that the unexpected changes during the dormant periods are due to changes in soil surface roughness caused by land management activities and show a relationship between changes in roughness and changes in optical thickness, which may be confusing the SMOS retrieval algorithm.

The COSMOS network will eventually consist of several hundred sensors throughout the United States that report kilometer-scale soil water content via measurement of the intensity of neutrons immediately above Earth's surface. We show that COSMOS sensors must be corrected for the effects of growing vegetation. Once this phenomenon is completely understood the COSMOS network could be a useful source of information for the validation of both soil moisture and vegetation products obtained from current and future microwave remote sensing satellites.

The European Space Agency's Soil Moisture and Ocean Salinity (SMOS) satellite mission provides microwave L-band measurements of vegetation optical thickness over the Earth. The optical thickness is related to water held in vegetation. The water content of crops varies over the growing season from a minimum at planting to a maximum during reproduction and back to a minimum during senescence. We found that in Iowa in 2010 the change in SMOS optical thickness over the growing season can be related to crop yields. However, there are inconsistencies in the optical thickness data, particularly high-frequency variation and unexpected changes during the non-growing season. We hypothesize that the unexpected changes during the dormant season are due to changes in soil surface roughness due to land management activities and show a relationship between changes in roughness and changes in optical thickness which may be confusing the SMOS retrieval algorithm.

This study investigates differences in the soil moisture product and brightness temperatures between 6 p.m. and 6 a.m. local solar time Soil Moisture Ocean Salinity (SMOS) passes for a region in the Midwest United States. This region has uniform land cover, consisting largely of maize and soybean row crops. The comparison was restricted to periods with no rainfall. There were 19 days available for analysis of the soil moisture product. It was found that there was a significant difference in the soil moisture product for all 19 days, with lower soil moisture for most mornings. The difference between the soil moisture products on some days exceeded the allowable error of 0.04 m3m-3. In-situ and model results indicate that there should be virtually no change in soil moisture between the evening and morning. In order to investigate this discrepancy, measured brightness temperature was converted to a polarization index (PI), and evening and morning values were compared. Investigation of the measured brightness temperature was limited to five days where a large range in incidence angles was available. Large differences between evening and morning passes were found for incidence angles less than 40° that could not be explained with radiative transfer theory but may be attributed to technical issues. There was also a difference in thePIvalues between the evening and morning passes for incidence angles greater than 40°. This can be caused by a decrease in soil moisture from evening to morning or could be attributed to an increase in the volumetric water content of the vegetation.

Evening (6pm) and morning (6am) SMOS observations are different. The greatest difference is for incidence angles less than approximately 30°, but this difference can not be explained with physical models. There is a smaller but significant difference in evening and morning SMOS observations for incidence angles greater than approximately 50°. Evening values of the polarization index are higher than morning values. This difference does have two plausible explanations: either soil moisture has decreased overnight; or vegetation water content has increased overnight.

NASA's Soil Moisture Active Passive satellite mission will use both an L-band radiometer and radar to produce global-scale measurements of soil moisture. L-band backscatter is also sensitive to the water content of vegetation. We found that a moderate dew increased the L-band backscatter of a soybean canopy by 1 dB. Dew thus has the potential to add error to satellite observations of soil moisture.

We have formulated a first-order tau-omega model. Compared to the commonly-used zero-order model, this model has four new terms that each represent a scattering mechanism. We found that the first mechanism, the scattering of emission from the vegetation into the upwelling beam, is the most significant. We also found that this term does not affect the overall soil moisture sensitivity such that the zero-order and first-order models at 1.4 GHz have essentially the same sensitivity to soil moisture. Adding this scattering mechanism will allow a new relationship between tau and the amount of vegetation within the canopy that does not increase tau as rapidly with vegetation biomass. The result may be a model that better matches experimental observations of the sensitivity of terrestrial microwave emission to soil moisture while still allowing the model to produce the correct 1.4 GHz brightness temperature.

Terrestrial microwave emission is sensitive to soil moisture. Soil moisture is an important yet unobserved reservoir of the hydrologic cycle linked to precipitation variability. Remote sensing satellites that observe terrestrial microwave emission have the potential to map the spatial and temporal variabilities of soil moisture on a global basis. Unfortunately, terrestrial microwave emission is also sensitive to water within the vegetation canopy, and the effect of free water residing on vegetation, either as intercepted precipitation or dew, is not clear. Current microwave emission models neglect the effect of free water. We found that the precipitation intercepted by a maize (corn) canopy increased its brightness temperature at 1.4 GHz. This effect is opposite that of dew: dew decreases the brightness temperature of maize at 1.4 GHz. The increase in brightness temperature due to the intercepted precipitation was only about 1 K for vertically polarized brightness temperature and about 3 K for horizontally polarized (H-pol) brightness temperature. It may be acceptable to neglect the effect of free water in microwave emission models. A more serious concern, however, is the underestimation, by current microwave emission models, of the sensitivity of the H-pol brightness temperature to soil moisture through maize. Understanding the physics associated with the effect of free water in vegetation on the emission, scattering, and attenuation of microwave radiation will lead to improved emission models, and potentially, models that correctly reproduce the sensitivity of the 1.4-GHz brightness temperature to soil moisture at high levels of biomass when vegetation effects are greatest.

A radiative transfer model is applied to data collected in a bare agricultural field in order to determine the usefulness of the method. Both measured and simulated soil moisture and soil temperature profiles are used as the inputs to the radiative transfer model. The simulation is carried out through a land surface model called The Atmosphere and Land-surface Exchange model (ALEX). It is observed that ALEX outputs are in close agreement with the measured soil moisture and soil temperature variations. The radiative transfer model can reproduce the brightness temperature variation in time when the soil moisture and temperature do not change abruptly. However, in the case of an abrupt change, the model fails to reproduce the brightness temperature profile. One possible reason is that the model assumes a uniform soil in terms of the volumetric water content while the abrupt change in the measured brightness temperature could be due to the existence of layers that have different water contents. We hypothesize that a model that can account for a non-uniform soil water content must be used.

We present the diurnal variation of vertical temperature differences measured within and beneath a maize canopy over the course of a growing season, and we analyze the implied temperature gradients in the context of microwave radiometry and soil moisture retrieval in particular. We find that the temperature differences can be as large as 9 K in magnitude within the vegetation canopy and as large as 10 K between the soil surface and a depth of 4.5 cm. Satellite overpass times at 1:30 A.M. and 1:30 P.M. occur close to when the magnitude of the temperature differences are largest. For 6 A.M. and 6 P.M. overpass times, temperature differences were smaller in magnitude at 6 P.M. This contradicts the widely held assumption that surface temperature gradients are more uniform at 6 A.M. than at 6 P.M.

Top-of-the-canopy infrared temperature measurements, and the mean of top-of-the-canopy and soil surface infrared temperature measurements, were equally appropriate definitions for vegetation canopy temperature when modeling the 1.4 GHz brightness of a maize canopy. The use of soil temperature as a surrogate for vegetation canopy temperature produced the largest error

We present the diurnal variation of vertical temperature gradients measured in a maize canopy over the course of a growing season in the context of the SMOS mission. For the 6 AM and 6 PM overpass times, gradients were actually smallest in magnitude at 6 PM. The mean, standard deviation, and extreme values of the temperature differences between the top of the canopy and the soil surface and between the soil surface and a depth of 4.5 cm at SMOS overpass times are reported.

The 1.4 GHz brightness of a field corn canopy can be predicted with the zero-order radiative transfer model only when the canopy is considered to be anisotropic.

At 1.4 GHz, dew on a corn canopy has the net effect of decreasing the brightness.

We investigate anisotropy in 1.4-GHz brightness induced by a field corn vegetation canopy. We find that both polarizations of brightness are isotropic in azimuth during most of the growing season. When the canopy is senescent, the brightness is a strong function of row direction. On the other hand, the 1.4-GHz brightness is anisotropic in elevation: an isotropic zero-order radiative transfer model could not reproduce the observed change in brightness with incidence angle. Significant scatter darkening was found. The consequence of unanticipated scatter darkening would be a wet bias in soil moisture retrievals through a combination of underestimation of soil brightness (at H-pol) and underestimation of vegetation biomass (at V-pol). A new zero-order parameterization was formulated by allowing the volume scattering coefficient to be a function of incidence angle and polarization. The small magnitude of the scattering coefficients allows the zero-order model to retain its limited physical significance.

The authors' point-scale Land Surface Process/Radiobrightness (LSP/R) model for a prairie grassland in the northern Great Plains was adapted to winter wheat-stubble within the region of the Southern Great Plains 1997 (SGP'97) Hydrology Experiment. The model maintains running estimates of near-surface soil moisture and stored water in soil and vegetation when forced by weather, and predicts the microwave brightness of the terrain. LSP/R model predictions were compared with the field observations recorded during SGP'97. The model captures canopy and soil temperatures very well, with the maximum mean and variance of the difference between the model and field temperatures being 1.06 K and 3.28 K/sup 2/, respectively. It yields reasonable predictions for the moisture in deeper layers of the soil, but its predictions for the moisture in the upper layers are low by /spl sim/2.3% by volume. These underpredictions of near-surface soil moisture result in higher H-pol brightnesses at 19 GHz than those observed.