The soil phase-transition temperature (PTT) is critical for accurately determining the soil freeze–thaw (FT) state, which is essential for developing surface FT state identification algorithms based on remote sensing technology and reliably validating FT products. In this study, we develop an innovative approach for determining soil PTT in the region of the dense soil moisture and soil temperature monitoring network in the Qinghai Lake Basin (QLB-NET). We employ four advanced models—polynomial regression (PR), random forest (RF), feedforward neural network (FNN), and transformer (XFMR) models—to establish relationships between soil PTT and various environmental factors for daily and site-specific soil PTT estimation. Soil PTT values derived from asigmoidcurve fitting (SCF) method that combines physical theory with empirical modeling are used in the models as the target data considering soil properties and topographic features as predictors. Additionally, we systematically evaluate the models’ performance and compare the results against those of the fixed 0°C threshold methods. The results indicate that most soil freezing temperatures range between -1°C and 0°C, and thawing temperatures range between -0.5°C and 0.5°C, with occasional freezing temperatures above 0°C. Model-derived soil PTT significantly outperforms the traditional 0°C threshold in terms of soil FT state classification accuracy, particularly during the frozen and FT transitional seasons. The XFMR model achieves superior performance in an internal comparative analysis, highlighting its particular effectiveness for soil PTT determination. These results provide valuable insights for improving remote sensing-based FT product validation and permafrost monitoring.

Due to the lack of in-situ measurements, direct validations can only be carried out at limited pixels and cannot be expanded to a regional or global scale. The three-cornered hat (TCH) method is essentially an optimal solution based on the numerical method, allowing an arbitrary number of datasets (at least three) to simultaneously participate in the calculation, with a tolerance to the error cross-correlation to a certain extent. However, in TCH, only a basic error model is adopted, which describes the products as mutually differing only in an additive random error. This violates our common recognition of the error model (i.e., affine error model) applied to describing products like soil moisture (SM). This work aims to improve the TCH method by updating the basic error model with the affine error model, and the reliability of the improved TCH method is evaluated using simulated soil moisture datasets and multisource soil moisture products against in-situ SM measurements.

The transition of soil freeze/thaw (FT) state plays a critical role in ecosystem, hydrological, and biogeochemical processes. However, obtaining representative soil FT state datasets with a long time sequence and fine spatiotemporal resolution remains challenging. Here, a decision-level spatiotemporal data fusion algorithm based on Convolutional Long Short-Term Memory networks (ConvLSTM) for soil FT product is proposed to enhance the enhanced L3 freeze/thaw product of the Soil Moisture Active Passive Mission (SMAP_E_FT). In the algorithm, the feature of long-time sequence from the Freeze/Thaw Earth System Data Record product (ESDR_FT) and the land surface dataset produced from the fifth generation ECMWF atmospheric reanalysis (ERA5-land) are sucked in. The results of the validation based on eight dense in-situ networks in Genhe, Maqu, Naqu, Ngari, Pali, Saihanba, Shandian river basin (SDR), and Tianjun show that the correlation coefficient between the fusion product (ConvLSTM_FT) and SMAP_E_FT is 0.9498, and the deviation is 0.0140, which maintains the classification accuracy. This study provides a possible way of temporally extending the SMAP_E_FT.

Mountains account for approximately 24% of the global land area and are crucial in regulating regional and global climate. However, the terrain not only introduces observation errors in the microwave remote sensing data acquisition process but also affects the microwave scattered radiation on the surface and the redistribution of hydrothermal energy, thereby influencing the accuracy of surface parameter inversion. To address this issue, this article incorporates roughness and vegetation models into the mountain radiation model to simulate the changes in surface microwave radiation in mountainous areas. By considering terrain parameters, soil moisture, and temperature, we can use Rugosity (RU) to accurately estimate the topographic effects on the brightness temperature of mountainous area. A higher RU value indicates a more substantial impact of soil moisture and temperature on the brightness temperature difference.

Thermokarst lake is a periglacial landfom that is widely distributed in permafrost area and seasonally frozen ground. Remote senisng data has been used to mapping the existance of thermokarst lakes. In this study, we used Sentinel-2 data, DEM data and the shape criteria to identify thermokarst lakes on Qinghai-Tibetan Plateau(QTP). The results were compared with former study and shows that the fomer study over estimated the area and number of thermokarst lakes in the study area.

The cross-domain recommendation (CDR) model addresses challenges such as data sparsity, the long tail distribution of user-item interactions, and the cold start of items or users. However, solely transferring domain-shared knowledge based on the co-occurrence patterns, without considering user preferences, leads to negative transfer in CDR. To overcome these limitations, we propose an advanced deep learning CDR model called the Domain Adversarial Deep Interest Network (DADIN) aims to facilitate smooth knowledge transfer from the source domain to the target domain and effectively alleviate negative transfer. Firstly, the joint distribution alignment of user preference in DADIN is realized by introducing a skip-connection-based domain agnostic layer, and then the domain classifier is artificially designed to distinguish between the information coming from the source domain or the target domain. Additionally, DADIN combines prediction loss, global domain confusion loss, and intra-class domain confusion losses through the Min-Max game and gradient reverse layer to achieve collaborative optimization. Two real-world experiments show the area under curve (AUC) of DADIN is 0.78 on the Huawei dataset, and it outperforms its competitors by 0.71% on the Amazon dataset, showcasing its state-of-the-art performance. Moreover, our ablation studies further demonstrate that domain adversarial technique increases the AUC by 2.34% on the Huawei dataset and 16.67% on the Amazon dataset, respectively.

Abstract Few‐shot learning enables machines to learn efficiently from limited labelled data. However, existing few‐shot learning methods may perform poorly when there is a lack of sufficient samples, and may encounter problems such as domain shift or overfitting when applied to new domains or tasks. To address the issues of poor fitting and insufficient generalization ability in new domains, a new method called triplet merged network with involution operators (TMNIO) is proposed. This method employs dual encoders that extract common and distinctive features from the prototype network, thereby enhancing the model's feature extraction capability. To further improve this ability, the traditional convolutional kernels are replaced with involution operators, which not only reduce the parameter count but also enlarge the receptive field to better extract local feature information. Additionally, this method employs a two‐stage training strategy, where triplet loss is used in the first stage to train the model and enhance its robustness and generalization ability. Extensive experiments on the miniImageNet, Omniglot, and Caltech‐UCSD Birds‐200 (CUB) datasets have shown that our proposed method achieved significant improvements in both training speed and accuracy, particularly on the miniImageNet dataset, where it achieved an outstanding 10% performance improvement.

Recommendation system algorithm based on multi-task learning (MTL) is the major method for Internet operators to understand users and predict their behaviors in the multi-behavior scenario of platform. Task correlation is an important consideration of MTL goals, traditional models use shared-bottom models and gating experts to realize shared representation learning and information differentiation. However, The relationship between real-world tasks is often more complex than existing methods do not handle properly sharing information. In this paper, we propose an Different Expression Parallel Heterogeneous Network (DEPHN) to model multiple tasks simultaneously. DEPHN constructs the experts at the bottom of the model by using different feature interaction methods to improve the generalization ability of the shared information flow. In view of the model's differentiating ability for different task information flows, DEPHN uses feature explicit mapping and virtual gradient coefficient for expert gating during the training process, and adaptively adjusts the learning intensity of the gated unit by considering the difference of gating values and task correlation. Extensive experiments on artificial and real-world datasets demonstrate that our proposed method can capture task correlation in complex situations and achieve better performance than baseline models.

There are two fundamental problems in applying deep learning/machine learning methods to disease classification tasks, one is the insufficient number and poor quality of training samples; another one is how to effectively fuse multiple source features and thus train robust classification models. To address these problems, inspired by the process of human learning knowledge, we propose the Feature-aware Fusion Correlation Neural Network (FaFCNN), which introduces a feature-aware interaction module and a feature alignment module based on domain adversarial learning. This is a general framework for disease classification, and FaFCNN improves the way existing methods obtain sample correlation features. The experimental results show that training using augmented features obtained by pre-training gradient boosting decision tree yields more performance gains than random-forest based methods. On the low-quality dataset with a large amount of missing data in our setup, FaFCNN obtains a consistently optimal performance compared to competitive baselines. In addition, extensive experiments demonstrate the robustness of the proposed method and the effectiveness of each component of the model.

Open intent detection is a significant problem in natural language understanding, which aims to identify the unseen open intent while ensuring known intent identification performance. However, current methods face two major challenges. Firstly, they struggle to learn friendly representations to detect the open intent with prior knowledge of only known intents. Secondly, there is a lack of an effective approach to obtaining specific and compact decision boundaries for known intents. To address these issues, this paper presents an original framework called DA-ADB, which successively learns distance-aware intent representations and adaptive decision boundaries for open intent detection. Specifically, we first leverage distance information to enhance the distinguishing capability of the intent representations. Then, we design a novel loss function to obtain appropriate decision boundaries by balancing both empirical and open space risks. Extensive experiments demonstrate the effectiveness of the proposed distance-aware and boundary learning strategies. Compared to state-of-the-art methods, our framework achieves substantial improvements on three benchmark datasets. Furthermore, it yields robust performance with varying proportions of labeled data and known categories. The full data and codes are available for use athttps://github.com/thuiar/TEXTOIR.

Multimodal intent recognition is a significant task for understanding human language in real-world multimodal scenes. Most existing intent recognition methods have limitations in leveraging the multimodal information due to the restrictions of the benchmark datasets with only text information. This paper introduces a novel dataset for multimodal intent recognition (MIntRec) to address this issue. It formulates coarse-grained and fine-grained intent taxonomies based on the data collected from the TV series Superstore. The dataset consists of 2,224 high-quality samples with text, video, and audio modalities and has multimodal annotations among twenty intent categories. Furthermore, we provide annotated bounding boxes of speakers in each video segment and achieve an automatic process for speaker annotation. MIntRec is helpful for researchers to mine relationships between different modalities to enhance the capability of intent recognition. We extract features from each modality and model cross-modal interactions by adapting three powerful multimodal fusion methods to build baselines. Extensive experiments show that employing the non-verbal modalities achieves substantial improvements compared with the text-only modality, demonstrating the effectiveness of using multimodal information for intent recognition. The gap between the best-performing methods and humans indicates the challenge and importance of this task for the community. The full dataset and codes are available for use at https://github.com/thuiar/MIntRec.

Here, four normalized difference water index (NDWI) variants, i.e., NDWI(860,970), NDWI(860,1240), NDWI(860,1640), and NDWI(1240,1640)are generated from the corn-oriented PROSAIL radiative transfer model. It is found that, instead of the linear relationship derived in previous studies, corn canopy water content (CWC) is best approximated as an exponential function of NDWI. Following the analysis of the PROSAIL-generated results, a newly optimized NDWI-based scheme is proposed for estimating corn CWC according to variations in the performance of the four NDWI variants under different CWC conditions. Validation results based on independent field data from the SMEX02, HiWATER2012, and Baoding2018 field experiments verify that this optimized NDWI-based corn CWC estimating scheme has a higher accuracy ($R = 0.87\,\,\pm \,\,0.03$, RMSE = 0.2068 ± 0.0145 kg/m2) than existing NDWI-based strategies for corn CWC retrieval. The feasibility of retrieving corn vegetation water content (VWC) based on the optimized NDWI-based scheme is also investigated, and the superiority of the optimized NDWI-based scheme for retrieving corn VWC is assessed. By comparing with four other NDWI-based corn VWC estimating methods, as well as the corn VWC parameterization scheme applied in the SMAP soil moisture algorithm, it is shown that our optimized NDWI-based scheme has the best VWC estimation accuracy, with the highest$R$of 0.89 ± 0.02 and the lowest RMSE of 0.7179 ± 0.0555 kg/m2.

The spatial heterogeneity of ground within a passive microwave pixel has a considerable effect on measured brightness temperature. In this paper, two types of soil moisture distribution patterns, the `center' pattern and the `matrix' pattern, were set up in a controlled field experiment to be observed by a microwave radiometer. Soil moisture was changed artificially during the experiment to form different soil moisture distributions within the footprint. The brightness temperatures of varying soil moisture conditions were simulated by the area weighted method (AWM) and antenna gain pattern weighted method(APWM). The effects of heterogeneous soil moisture on observed brightness temperatures were explored by comparing the brightness temperatures between measured and simulated values. It has shown that the simulated brightness temperatures matched well with observed values for the `center' pattern using AWM. In the `matrix' pattern, however, the brightness temperatures calculated using APWM were more coincident with the observations comparing to values calculated by AWM.

The application of passive microwave soil moisture products is limited by the poor spatial resolution. In this study, a spatially-based regression method was tested using the PLMR(Polarimetric L-band Multibeam Radiometer) observation data acquired in the Heihe Watershed Allied Telemetry Experimental Research (HiWATER). The land surface temperature data were used as the auxiliary data to perform downscaling soil moisture and brightness temperature. The root-mean-square errors (RMSEs) of brightness temperature and soil moisture are 5.5 K and 0.021 cm3/cm3between the calculated values and refenced values, respectively. This method provided a potential way for the downscaling of passive microwave brightness temperature and soil moisture.

The relief of land surface will bring error into the retrieval of land surface parameters, e.g. soil moisture, snow water equivalent, etc., using passive microwave remote sensing, especially in mountainous areas. However, until recently, the topography effects on the microwave radiation of mountainous areas didn't arouse widely concern. Based on the simulation scheme of former researches, we simulated the microwave radiation of a 40 km by 40 km square pixel, of which the surface temperature and soil moisture were inhomogeneous. The simulated brightness temperature shows noticeable angler dependence with respect to both the azimuth and elevation angle.

Vegetation water content (VWC) is a key variable in land-atmosphere interactions and plays an important role in agriculture, climate and hydrology. Based on the first-order scattering model, simulation database of corn backscattering coefficients at L-band was established. The simulations were used to train an artificial neural network (ANN) to establish an inverse model for corn VWC estimation during corn growth periods. The inverse accuracy of the trained ANN was evaluated using ground corn samplings and radar data acquired by the Passive and Active L- and S-band (PALS) airborne microwave sensor during the Soil Moisture Experiments in 2002 (SMEX02). Moreover, the corn VWC inversion results were compared to those obtained from an empirical method using the radar vegetation index (RVI). Result showed that the ANN method is superior to the RVI method and capable of estimating corn VWC with a correlation coefficient (R) of 0.7987, a root mean square error (RMSE) of 0.6033 kg/m2and a mean absolute relative error (MARE) of 12.00%.

The matrix doubling (MD) model is a numerical solution to the radiative transfer equation. It can achieve better accuracy in simulating microwave signals from vegetated terrain by considering multiple-scattering effects. However, it is difficult to apply the MD model to retrieving work due to its high complexity. This letter presents a case study performed on corn to demonstrate a multiangular (5°-65°), multiband (1.4/6.925/10.65 GHz) microwave emission model considering multiple-scattering effects by parameterizing the MD model. The simulated emissivity differences between the theoretical model and parameterized model are small. The mean absolute percent errors are all less than 1%, and the root mean square errors (RMSEs) are all within the range of 10-3. Validations using airborne polarimetric L-band microwave radiometer data and ground-based trunk-mounted multifrequency microwave radiometer data indicate that the parameterized model achieves good accuracy with overall RMSEs within 8K at all three bands.

According to the microwave radiation characteristics, this paper introduced a new frozen soil dielectric model to calculate the dielectric constant of frozen and thawed soil based on the Helsinki University of Technology (HUT) microwave snow emission model. The Advanced Integrated Emission Model (AIEM) was used to calculate surface emissivity. The multi-frequency microwave radiation model and soil freeze/thaw discriminant algorithm were improved. The classification accuracies of original and improved soil freeze/thaw discriminant algorithms were validated using AMSR2 Level 3 daily gridded 0.25° brightness temperature products and the measured values obtained by ground-based microwave radiometer. The results showed that compared to the original discriminant algorithm, the frozen soil classification accuracy of the improved discriminant algorithm was effectively improved and the overall classification accuracy reached 82%. It turned out to be a comparatively reliable mode of discrimination.

The component soil moisture within one mixed-pixel is important for predicting soil moisture contents in eco-hydrological research in arid and semi-arid regions. The multi-angular observations of passive microwave remote sensing provide the possibility to derive mixed-pixel component soil moisture. In this study, a field experiment was carried out using a truck-mounted multi-frequency microwave radiometer (TMMR). Because on the difference of radiative characteristics in parallel-to-row and perpendicular-to-row directions, an algorithm for estimating the component soil moisture within a footprint was developed. Taking the row-structure effect into consideration, the component soil moisture could be estimated based on the tau-omega model. It has shown that the RMSE decreased when comparing the component soil moisture with the corresponding soil moisture measurements in bare surface.

Ground-based microwave radiometric measurements were carried out in 2016 by using a multi-frequency microwave radiometer at L, C and X bands (1.4, 6.925 and 10.65 GHz). The aim of the experiments was to explore multi-frequency microwave emission characteristics of the soil freeze-thaw process for model and algorithm development for the future Water Cycle Observation Mission (WCOM). Measurements were carried out on pastureland in Chengde, Hebei Province, which belongs to seasonally frozen ground of China. Soil temperature and soil moisture profiles, the frost depth, and meteorological observations were synchronously collected. It has been found that microwave radiation has different responses to soil freezing and thawing process at different frequencies.

Atmospheric transmittance is an important factor for atmospheric correction in the inversion of land surface parameters. Under nonprecipitating conditions, microwave atmospheric transmittance in the X-, Ku-, and Ka-bands is mainly determined by oxygen, water vapor, and cloud liquid water content. In this letter, radiosoundings from 119 stations in China, performed twice a day from January 2011 to July 2014, were used in the Salonen-Uppala cloud detection algorithm to distinguish cloud layers from layered atmospheric profiles and to estimate the cloud liquid water content therein. The resulting atmospheric transmittances at frequencies of the Advanced Microwave Scanning Radiometer-Earth Observing System over China were estimated using Liebe's millimeter-wave propagation model and Mie theory. Atmospheric transmittance maps were obtained by interpolating the results from individual sites, driving a climatological database for over China, which may be used to correct for atmospheric influence in surface parameter retrievals. The simulated transmittances were validated through a series of on-site field experiments in the North of China. We compared simulated atmospheric brightness temperature with measurements performed in situ using a ground-based radiometer system. The correlation coefficients between the measured and simulated values were 0.97, 0.99, and 0.98, in the X-, Ku-, and Ka-bands, respectively, with a root-mean-square error of 0.6, 1.6, and 9.5 K, respectively.

Passive microwave remote sensing has all-weather work capabilities, but atmospheric media have different influences on satellite microwave brightness temperature under different atmospheric conditions and environments. In order to clarify atmospheric influences on Advanced Microwave Scanning Radiometer-Earth Observing System (AMSR-E), atmospheric radiation were simulated based on AMSR-E configuration under clear sky and cloudy conditions, by using radiative transfer model and atmospheric conditions data. Results showed that atmospheric water vapor was the major factor for atmospheric radiation under clear sky condition. Atmospheric transmittances were almost above 0.98 at AMSR-E's low frequencies (<18.7GHz) and the microwave brightness temperature changes caused by atmosphere can be ignored in clear sky condition. Atmospheric transmittances at 36.5GHz and 89GHz were 0.896 and 0.756 respectively. The effects of atmospheric water vapor needed to be corrected when using microwave high-frequency channels to inverse land surface parameters in clear sky condition. But under cloud covered conditions, cloud liquid water was the key factor to cause atmospheric radiation. When sky was covered by typical stratus cloud, atmospheric transmittances at 10.7GHz, 18.7GHz and 36.5GHz were 0.942, 0.828 and 0.605 respectively. Comparing with the clear sky condition, the down-welling atmospheric radiation caused by cloud liquid water increased up to 75.365K at 36.5GHz. It showed that the atmospheric correction under clouds covered condition was the primary work to improve the accuracy of land surface parameters inversion of passive microwave remote sensing. The results also provided the basis for microwave atmospheric correction algorithm development. Finally, the atmospheric sounding data was utilized to calculate the atmospheric transmittance of Hailaer Region, Inner Mongolia province, China, in July 2013. The results indicated that atmospheric transmittances were close to 1 at C-band and X-band. 89GHz was greatly influenced by water vapor and its atmospheric transmittance was not more than 0.7. Atmospheric transmittances in Hailaer Region had a relatively stable value at low frequencies(<18.7GHz) in summer, but had about 0.1 fluctuations with the local water vapor changes at high frequencies.

Passive microwave remote sensing provide soil moisture product. The problem of validation of soil moisture products is that the true soil moisture of such a large pixel is difficult to estimate. This paper is a preliminary research on how to measure the true soil moisture of a 30m×30m quadrat. The results suggest that 20 measurement in such a quadrat should gave a reliable estimation of the soil moisture of the quadrat. This could be used as a thumb rule in designing the validation plan of remote sensing soil moisture product. And also could be used to estimate the soil moisture distribution within a space-born microwave radiometer footprint.

Passive microwave remote sensing has demonstrated the potential for capturing the high temporal variability of the near-surface soil moisture, however the use of these data is limited by the poor spatial resolution. We compared two different downscaling methods using soil evaporative efficiency derived from Moderate-resolution Imaging Spectroradiometer (MODIS) to disaggregate AMSR-2 soil moisture product. Both methods used information from MODIS to obtain the distributed soil moisture map at regional scales and the results showed reasonable agreement with ground-based soil moisture observations. For Merlin downscaling results, the correlation coefficient and RMSE with ground measured data are 0.74 and 3.21%, and for UCLA downscaling results, the correlation coefficient and RMSE with ground measured data are 0.76 and 3.81%, respectively.

The dielectric property of vegetation has a considerable effect on the characteristics of the microwave radiation of vegetation. In frozen environments, when the temperature is colder than normal, changes such as increased soluble sugar and decreased moisture content (MC) can occur in the vegetation. The dielectric property of vegetation, which is almost entirely controlled by its free and bound water content, will also change. To characterize the dielectric behavior of vegetation in frozen regions, a sensitive experiment was conducted on holly leaves with a high-performance coaxial probe over a frequency range from 0.5 to 40 GHz and a temperature range from 0°C to -20°C. Based on the measurements and the physical properties of the constituent substances of vegetation, a semiempirical dielectric model for holly leaves in low temperature environments was developed. In this model, a decrease in MC, which causes a reduction in the complex permittivity, was described as an increase in the ice content. The complex permittivity of bound water was measured using a saturated sucrose solution at -6.5°C. The research will provide a reference for the dielectric property study of the vegetation in frozen environments.

To retrieve the soil salinity by microwave remote sensing, we must clarify the relation of dielectric properties of saline soils with soil salinity. The objective of this paper was to determine how dielectric properties are affected by soil salinity in the remote sensing range and develop an improved model applying for the saline soil. Laboratory measurements of dielectric constant are made with a microwave vector network analyzer using soil mixture samples of various soil moistures and salinities prepared artificially from natural soils. The results confirmed that the real part is strongly affected by soil moisture, whereas the imaginary part depends on both the soil moisture and salinity, particularly at lower frequencies (1-6 GHz). Thus, as a key factor, soil salinity is introduced into the expression of the imaginary part of dielectric constant in the Dobson semiempirical dielectric mixing model, combining the dielectric model for saline water and the impact of electrical conductivity of soil solution. The improved model yields results, which are in good agreement with the laboratory measurements, the slopes of fitting curves between measurements and simulation nearly being equal to 1, and coefficient R2are higher than 0.89. In addition, the improve model is independent of soil-texture parameters. However, it should be noted that the improved model can be applied only to the C-band remote sensing.

Soil moisture is an important parameter in many fields. Since the dielectric constant of soil is directly related with its moisture content, many soil dielectric constant models have been established and used in the application of soil moisture inversion. As an effective composition of soil, organic matter could increase the adsorption of soil particles and affect the dielectric constant. However, due to its little content, it was seldom considered in soil moisture inversion and brightness simulation. In this study, a semi-empirical organic dielectric model was used in the forward simulation of brightness temperature in Genhe River basin. The results show that it has a higher accuracy about 1.6k~2.4k than using TMD model at C-band and X-band.

In this study, winter wheat water content was retrieved using a parameterized relationship between the winter wheat water content and its structure parameters, i.e. radius, length and thickness of the leaves. The parameterized relationship was constructed from a large quantity of simulated attenuation cross section (Qe) of the winter wheat leaves based on the Generalized Rayleigh-Gans approximation (GRG) and the Mätzler vegetation dielectric constant model. The study was conducted at 6.925GHz and under the assumption that there was no significant polarization difference between the vegetation signals. This new method were applied to ground-based brightness temperature observations to retrieve winter wheat water contents, which were further compared with the corresponding filed measurements. Results show that the retrieved winter wheat water contents are consisted with the field measurements. It proves that the winter wheat inversion method proposed in this study is feasible. Moreover, the method is promising to apply to the satellite observations.

A useful indicator to evaluate the soil freeze-thaw intensity is the amount of phase transition water content (PTWC) in soil pores. In this research, a power function relation between soil phase transition water content (PTWC) and the variation rate of land surface temperature (VTS) was found by analyzing of ground measured soil moisture and temperature data obtained in the Tibet plateau during the winter of 2012. Then a downscaling approach combining MODIS VTsand AMSR2 products was employed to retrieve high resolution PTWC. The downscaled result was tested using in situ observations from the CTP-SMTMN network and found that it quite followed the trend of ground data with a RMSE of 0.0034 (m3/m3) and MAE of 0.0025 (m3/m3). The comparisons indicate that PTWC-VTSmodel has combined the advantage of microwave remote sensing and optical remote sensing; it has a high precision and can generate PTWC in small scale.

This letter introduces the ecohydrological wireless sensor network (EHWSN), which we have installed in the middle reach of the Heihe River Basin. The EHWSN has two primary objectives: the first objective is to capture the multiscale spatial variations and temporal dynamics of soil moisture, soil temperature, and land surface temperature in the heterogeneous farmland; and the second objective is to provide a remote-sensing ground-truth estimate with an approximate kilometer pixel scale using spatial upscaling. This ground truth can be used for validation and evaluation of remote-sensing products. The EHWSN integrates distributed observation nodes to achieve an automated, intelligent, and remote-controllable network that provides superior integrated, standardized, and automated observation capabilities for hydrological and ecological processes research at the basin scale.

Freeze-thaw erosion is the third largest soil erosion type after water erosion and wind erosion, which is a serious threat to agricultural land and various buildings, especially for water projects. In this paper, a new method based on the passive microwave remote sensing technique was proposed to classify and assess the freeze-thaw erosion. The core of this new method is at two important indices: the freeze-thaw cycling days per year, and the phase transition water content per day. The first index can be used to determine the freeze-thaw erosion region and the second index can be used to evaluate the degree of freeze-thaw erosion. The application of this new method in China shows good result. It indicates that the freeze-thaw erosion regions in China are mainly distributed in Tibet, Mongolia and the province of Qinghai, Xinjiang, Gansu, Sichuan and Heilongjiang. Furthermore, the comparably serious freeze-thaw erosion region is located in Tibet Plateau.

Dielectric constant is an important parameter in microwave remote sensing. The microwave scattering/radiation signal of vegetation is closely related to its dielectric constant. Many related models have been established by now. However, most of them can only be used in room temperature. Therefore, it brings errors in the research of vegetation in frozen environment. In this study, a new dielectric model, which can be used at frequencies ranged from 3GHz to 40GHz under temperatures between -20°C and -4°C, has been established. It was developed based on Debye-Cole dual-dispersion model. The validation shows it has an acceptable precision.

In this study, the dielectric constants of 12 types of soil with different organic matter content were measured using the coaxial probe method by network analyzer (0.5-40 GHz) at room temperature (approx. 23°C). The observed dielectric constant increases only slowly with soil volumetric water content up to a transition point. Beyond the transition point, it increases rapidly with volumetric water content. It was found that the value of the transition point was higher and the observed dielectric constant was lower at the same soil volumetric water content and frequency for soil with higher organic matter content. A simple semi-empirical model was proposed to describe the dielectric behavior of soil with organic matter. This model was developed based on the refractive mixing dielectric model (RMDM).

A new dielectric model for vegetation in frozen environment based on the Debye-Cole dual-dispersion model was already developed in part I. This model can be used at a wide frequency range (0.5GHz - 40GHz) and even applicable for negative temperatures reached -20°C. In this paper, a matrix-doubling microwave emission model was used to evaluate vegetation effects in a frozen environment at 6.925, 10.65, 18.7 and 36.5GHz (V and H polarization). To verify the new developed dielectric model, a kind of young tree named Populus tomentosas was measured based on the Truck-mounted Multi-frequency Microwave Radiometer in December of 2012. In the experiment, the ground was irrigated to get rid of the soil signals. Also, the row-structure's effect on the trees can be eliminated when water covered the whole ground surface. Comparisons and analysis between model simulations and field measurements showed the dielectric model can be applied to microwave emission model as input data. Furthermore, the characteristics of microwave radiation of vegetation in frozen environment were evaluated and how the vegetation dielectric constant affected the electric field and physical property of vegetation layer was explained.

Penetrability is one of the greatest advantages of microwave remote sensing over other remote sensing techniques. Estimating the sensing depth of passive microwave remote sensing is meaningful for simulation of satellite signals and validation of land surface parameters. In this paper, microwave radiation response depth (MRRD) was proposed to describe the thickness of the soil layer, within which variations in soil moisture can significantly affect the emitted radiation. Then, a statistical model for estimating MRRD was developed by regression analysis. This model can estimate the MRRD using four parameters, which are soil moisture, soil temperature, frequency, and soil specific surface area. For validation, a controlled field experiment was performed using a truck-mounted multi-frequency microwave radiometer (TMMR) at Baoding, Hebei province, China. The accuracy of the statistical model, in terms of root mean square error (rmse), was approximately 0.54 cm for the available experimental data at the frequency of 6.925, 10.65, and 18.7 GHz. Finally, the statistical model was applied into soil moisture validation. Taking the Global Land Data Assimilation Systems (GLDAS) product as the measurements, we validated the soil moisture retrieved from advanced microwave scanning radiometer-earth observing system (AMSR-E) data. The measured soil moisture obtained from an empirical depth and MRRD were both used to comparison. The rmse between soil moisture retrieved and measured is 0.060 and 0.047 cm3/cm3for empirical depth and MRRD based method, respectively. These results have shown that it is more reasonable to considered MRRD than an empirical depth in soil moisture validation.

Phase transition water content (PTWC) is an important indicator of the intensity of soil freezing-thawing and is the crucial parameter that influences land surface energy balancing. In this paper, a PTWC retrieval algorithm was used to estimate the PTWC of pixel-scale. Through the field experiment at North China Plain, the actual PTWC is obtained to evaluate the accuracy of inversion. The validation result showed that the algorithm based on AMSR-E can provide reliable values of PTWC in pixel-scale. The RMSE of all sampling points is 0.008v/v.

A bistatic scatter model developed based on the Michigan grassland coherent model was used to simulate bistatic scattering coefficients from wheat field at L and C band. And the wheat field emissivities at V and H polarization were also obtained by integrating the bistatic coefficients over all scattering angles above ground. An experiment was carried out over a flat agriculture area located at Baoding, in the Hebei province of China. The plant and soil parameters were collected over a growing season, and the emissivities were also obtained using the ground-based microwave radiometer at the C-, X-, Ku-, and Ka- bands, in H and V polarizations. The radiometer model values are in good agreement with the measurements at some full growth stages.

Penetration depth of passive microwave remote sensing indicates the area below the land surface where singles come from. A reliable penetration depth is helpful not only in understanding the mechanism of microwave remote sensing but also in validating the retrieved results of land parameters. In this paper, we defined that the penetration depth in passive microwave remote sensing is the soil depth above which contributes more than 99% power of the total signal. Based on a three-layer non-coherent model (air-soil-aluminum), sensitive analysis was conducted to find out parameters which have more effect on penetration depth. Then three parameters including soil moisture, frequency and soil temperature were selected to build a simple statistic model to calculate the penetration depth of passive microwave remote sensing. To validate the statistic model, two controlled experiments were designed and carried out using a Truck-mounted Multi-frequency Microwave Radiometer (TMMR). It was shown that the penetration depth calculated from field experiment data coincided well with the statistic model.

The technique of microwave remote sensing has been used to monitor the soil frozen/thawed status for many years. The dielectric constant of frozen soil is relatively lower than that of unfrozen soil, so that microwave could penetrate deeper into frozen soil. However, we are still lack of the knowledge of the Microwave Penetration Depth (MPD) of frozen soil. In this study, a noncoherent microwave radiation was used to find the factors that influence the MPD of frozen soil and then validated by experiments. The results showed that the frequency of microwave, the temperature of frozen soil and the soil texture are the main factors that determine the MPD of frozen soil. An empirical model that estimates the MPD of frozen soil was proposed based on the simulation data.

Spectral Gradient (SG) is one of indictors that had been used to classify prairie soil as either frozen or thawed. Researchers suggested that negative spectral gradients are caused by volume scattering darkening within the frozen soil. A radiative transfer version of first order that considered volume scattering was used to simulate the brightness temperature of frozen soil. On the basis of sensitivity analysis, the prediction of this model was compared with ground experimental measurements. The results show that the volume scattering effect should be considered when modeling and measured BT of frozen soil especially when the temperature of ground is low.

Simulation of brightness temperature and related snow parameters is essential to understand the microwave emission property and its evolution with change of the snow soil system status. In this paper, a typical thin snow pack on North China Plain is measured on Nov 13-16th, 2009 at Luancheng test site HUT (Helsinki University of Technology) wet snow emission model is used to predict the brightness temperatures at 10.65, 18.7 and 36.5 GHz. A physically-based snow process model, SNTHERM (SNow THERmal Model), is applied to simulate the snow melting process. The measured snow density and grain size is compared with SNTHERM prediction and HUT inputs. Results show that the application of snow emission model and process model can explain the variation trend of wet snow emission properties well.

Vegetation water content (VWC) is a useful parameter in agriculture, forestry and hydrology studies. It is particularly valuable in accounting for vegetation effects in retrieving soil moisture using microwave radiometers. Microwave vegetation indices (MVIs) reflect information of the whole vegetation canopy. They may provide a mean for estimating VWC. In this study, a methodology for retrieving VWC using MVIs is presented. Coefficients of the relationship were found to be dependent only on a vegetation structure parameter. The methodology was tested with brightness temperature observations at C and X bands collected over a growing season of cotton. It was found that results compared well with field observations of VWC measured during the early growing season. The methodology should be useful for vegetation monitoring and soil moisture retrieval over low vegetated areas.

To investigate relief effects on microwave radiation, it is essential to experiment, based on track-mounted microwave radiometer. There are four relief factors affecting microwave radiation features in this study we have researched, which are hill slopes, hill elevation, hill aspects, and hill shadows. To compare with relief effect simulation, we built relief landscape in the field experiment to validate microwave radiation of hill-scale topography bias resulted from some of relief factors. In the final analysis, through the relief experiment observed results we consider hill-scale topography dose have influence on microwave radiation, and it can not be ignored in the retrieval of surface parameters.

Spectral gradient is one of the indicators of soil freeze and thaw. A negative spectral gradient was a sign of frozen soil. A truck-mounted microwave radiometer observed the frozen ground in two experiments. However, the measured spectral gradient was positive when the surface is frozen. From experiment measured and model simulated result, the relationship between the amount of unfrozen water of frozen soil and SG was found. It indicates that the amount of unfrozen water is a very important factor that determines the microwave radiation of frozen soil.

Ground surface freeze-thaw cycles caused by changes in solar radiation have a great impact on soil-air water heat exchanges due to the phase transition of pore water. This influence should not be ignored in the land surface process and global environment change studies because of its large extent and the rapid changes in daily and seasonal frozen ground. The key index for evaluating the influence intensity is the content of water-ice phase transition in soil pores at the ground surface. In this paper, a data set was generated by observing field experiments and physical model simulations based on the configuration of the Advanced Microwave Scanning Radiometer-EOS (AMSR-E). The results showed that microwave radiation from freezing/thawing soil has an obvious correlation to the phase transition process of soil water. A large change in soil surface emissivity was shown after the freezing of soil. The magnitude of the difference in emissivity change is strongly related to the amount of water-ice phase transition. It can be shown that the higher the phase transition water content (PTWC), the greater the emissivity difference, and the higher the frequency, the smaller the emissivity difference. Based on an analysis of a large amount of random simulation data, an interesting characteristic was found, in that the emissivity difference in vertical polarization at each frequency is nearly proportional to the phase transition water content. Thus, a ratio index called Quasi-emissivity (Qe) was developed to eliminate temperature effects during retrieval. Using these clear rules, a physical statistical algorithm was put forth to estimate the phase transition water content. Finally, the inferred results by ground-based radiometer observation were compared with the ground truth. A satisfying agreement was achieved with a root mean square error of 0.0265 (v/v). This indicated that the microwave radiometer has a great potential in the measurement of PTWC.

Snow plays an important role in meteorological and hydrological studies, so it makes sense to accurately predict the process of snow and the amount of snow. Exactly modeling snow properties is an important process for the combined snow process model and microwave model to simulate the amount of snow. In this paper, a mass and energy balance computer model-snow thermal model (SNTHERM.89) is used to simulate the snow properties combined with experimental data measured in Binggou basin, an alpine catchment in Gansu province, china during March 11th and April 7th in 2008. SNTHERM can simulate the snow properties well. In an attempt to make sure that the data for the input is with highest degree of confidence when some measurements are missing, sensitivity analysis of snow properties to forcing data was conducted. Through evaluating the sensitivity of SNTHERM to forcing data, a better understanding of the model and prediction can be obtained.

As the global warming intensifies, the environment changes in cold land receive more attention. In this paper, a combined microwave emission model is established for cold land researches. Through field observation experiment, the b-factor of winter wheat during winter is obtained to simulate radiation accurately from this typical ground object in China. Furthermore, the impacts of snow and vegetation cover on frozen soil radiation are investigated by sensitivity analysis.

Interference effect happens in layered medium. The brightness temperature oscillation has been observed during the freezing process of over-saturated soil, which could be explained by interference effect and a three layer coherent model. The modeled BT is qualitatively in consistent with the measurement. It is shown that the interference must be considered when measuring frozen soil with ground based microwave radiometer especially when using the frequency is low.

A multi-frequency microwave radiometer has been built to study the microwave remote sensing of soil, snow and vegetation cover. Its accuracy and stability has been tested. This article described the performance of its design and the performance of its components. Two absolute calibration methods and a relative calibration method and the calibration results are introduced. The theoretical and practical accuracy of the measured brightness temperature is in the range of 0.3 K-1.0 K. It is proved to be reliable and helpful in relative microwave remote sensing researches.