Soil Moisture Retrieval Using Microwave Remote Sensing Data and a Deep Belief Network in the Naqu Region of the Tibetan Plateau

Soil moisture plays an important role in the land surface model. In this paper, a method of using VV polarization Sentinel-1 SAR and Landsat optical data to retrieve soil moisture data was proposed by combining the water cloud model (WCM) and the deep belief network (DBN). Since the simple combination of training data in the neural network cannot effectively improve the accuracy of the soil moisture inversion results, a WCM physical model was used to eliminate the effect of vegetation cover on the ground backscatter, in order to obtain the bare soil backscatter coefficient. This improved the correlation of ground soil backscatter characteristics with soil moisture. A DBN soil moisture inversion model based on the bare soil backscatter coefficients as the foundation training data combined with radar incidence angle and terrain factors obtained good inversion results. Studies in the Naqu area of the Tibetan Plateau showed that vegetation cover had a significant effect on the soil moisture, and the goodness of fit (R2) between the backscatter coefficient and soil moisture before and after the elimination of vegetation cover was 0.38 and 0.50, respectively. The correlation between the backscatter coefficient and the soil moisture was improved after eliminating the vegetation cover. The inversion results of the DBN soil moisture model were further improved through iterative parameters. The model prediction reached its highest level of accuracy when the restricted Boltzmann machine (RBM) was set to seven layers, the bias and R were 0.007 and 0.88, respectively. Ten-fold cross-validation showed that the DBN soil moisture model performed stably with different data. The prediction was further improved when the bare soil backscatter coefficient was used as the training data. The mean values of the root mean square error (RMSE), the inequality coefficient (TIC), and the mean absolute percent error (MAPE) were 0.023, 0.09, and 11.13, respectively.

Assessment of EnKF Data Assimilation of Satellite Derived Soil Moisture Over the Indian Domain with the Noah Land Surface Model

Land Surface Models (LSMs) are typically forced with observed precipitation and surface meteorology and hence the soil moisture estimates obtained from LSM do not reflect the contribution of irrigation to the soil moisture estimates. However, the satellite retrievals of soil moisture estimates do register the signature of the irrigation effects. It is suggested that the soil moisture estimates obtained from LSM may reflect the role of irrigation if they are assimilated with soil moisture estimated from satellites. The present study evaluates the improvement of soil moisture estimates obtained from Noah LSM by ingesting them with the satellite derived Advanced Scatterometer (ASCAT) soil moisture retrievals over the Indian domain for the year 2012. The above ingesting of soil moisture estimates is performed using the Land Information System (LIS). The improved soil moisture estimates are validated with the in-situ India Meteorological Department (IMD) soil moisture observations and also with the high-resolution Indian Monsoon Data Assimilation and Analysis (IMDAA) regional reanalysis data. The percentage of grid points over the Indian domain where the improvement parameter shows positive values are 59.14% (winter), 69.17% (pre-monsoon), 43.59% (monsoon), and 77.53% (post-monsoon). Furthermore, the forecast impact parameter also indicates the positive impact of data assimilation. Also, 12 of the 22 stations show reduced RMSE soil moisture error after data assimilation is performed while only 6 of the 22 stations show higher correlation coefficient in soil moisture without data assimilation, when validated with the in-situ IMD soil moisture observations. The study has also evaluated the irrigation impact of ASCAT in the assimilated soil moisture using triple collocation (TC) method. For the TC analysis, the model based Global Land Data Assimilation System (GLDAS)Catchment Land Surface Model (CLSM), and MERRA (Modern-Era Retrospective analysis for Research and Applications) Land data set together with soil moisture model outputs with and without ASCAT assimilation are used to calculate the error and correlation coefficient of each of the two set of triplets. The results of the TC analysis further conclusively shows the positive impact of irrigation effects in the ASCAT assimilated soil moisture model output.

CAS-LSM Datasets for the CMIP6 Land Surface Snow and Soil Moisture Model Intercomparison Project

The datasets of the five Land-offline Model Intercomparison Project (LMIP) experiments using the Chinese Academy of Sciences Land Surface Model (CAS-LSM) of CAS Flexible Global-Ocean-Atmosphere-Land System Model Grid-point version 3 (CAS FGOALS-g3) are presented in this study. These experiments were forced by five global meteorological forcing datasets, which contributed to the framework of the Land Surface Snow and Soil Moisture Model Intercomparison Project (LS3MIP) of CMIP6. These datasets have been released on the Earth System Grid Federation node. In this paper, the basic descriptions of the CAS-LSM and the five LMIP experiments are shown. The performance of the soil moisture, snow, and land-atmosphere energy fluxes was preliminarily validated using satellite-based observations. Results show that their mean states, spatial patterns, and seasonal variations can be reproduced well by the five LMIP simulations. It suggests that these datasets can be used to investigate the evolutionary mechanisms of the global water and energy cycles during the past century.

Evaluation of the soil moisture model with data assimilation by the triple collocation method

This paper deals with a nonlinear soil moisture transport problem, solved with addition of satellite observed soil moisture. The satellite data are assimilated into the model using Newtonian nudging method. Evaluation is done by the triple collocation method, which involves three independent data sources: model results, ground stations and ERA5 climatic data. The results testify that model results are nearly as accurate as the ground station measurements.

The value of ASCAT soil moisture and MODIS snow cover data for calibrating a conceptual hydrologic model

Recent advances in soil moisture remote sensing have produced satellite datasets with improved soil moisture mapping under vegetation and with higher spatial and temporal resolutions. In this study, we evaluate the potential of a new, experimental version of the ASCAT Soil Water Index dataset for multiple objective calibration of a conceptual hydrologic model. The analysis is performed in 213 catchments in Austria for the period 2000–2014. An HBV type hydrologic model is calibrated to runoff data, ASCAT soil moisture data, and MODIS snow cover data for various calibration variants. Results show that the inclusion of soil moisture data in the calibration mainly improves the soil moisture simulations; the inclusion of snow data mainly improves the snow simulations; and including both of them improves both soil moisture and snow simulations to almost the same extent. The snow data are more efficient in improving snow simulations than the soil moisture data are in improving soil moisture simulations. The improvements of both runoff and soil moisture model efficiencies are larger in low elevation and agricultural catchments than in others. The calibrated snow-related parameters are strongly affected by including snow data, and to a lesser extent by soil moisture data, while the soil-related parameters are only affected by the inclusion of soil moisture data.

Improving the understanding of recharge in a basalt aquifer based on a soil moisture model, water levels and climatic data

<p>Knowledge of recharge processes in groundwater resource areas is of great importance for developing sustainable water management plans. In an effort to enhance the understanding of recharge in a basalt aquifer, a national water balance soil moisture model was compared with the response in water tables in multiple private pumping bores across the Tamborine Mountain plateau located in South East Queensland, Australia. The water levels in the pumping bores were influenced by the everyday use of the bores, which are utilised for household supply, stock watering, garden watering and irrigation. In each bore, the pumping response was identified and filtered out before being compared to the soil moisture model results. The soil moisture model (AWRA-L Australian Water Resource Assessment Landscape) includes results of surface runoff, soil moisture, evapotranspiration and deep drainage, to a depth of 6 m. The simulated soil moisture levels in the rootzone (rootzone defined as depth between 0 - 1 m), showed a similar hydrographic response following rain events to that observed in water levels in the aquifer. The response in the aquifer compared to the soil moisture showed some of the deeper bores had a lag effect and furthermore, the response also showed dependency on the soil moisture level (%) and on the size/duration of the rain event. It was observed that the simulated deep drainage (recharge) did not correlate to the observed changes in water tables. The soil moisture model simulated a nearly constant deep drainage (recharge) of 0.05±0.01mm a day, whereas the bores showed large increases in water table in response to rainfall events. Previous studies in the area based on the chloride mass balance approach have estimated that the annual deep drainage volume was an average of 30% of annual rainfall, while the soil moisture model approach has simulated an annual deep drainage volume of 1.2 – 1.7% of the total annual rainfall. While these results show that there are shortcomings related to applying the soil moisture model to estimate aquifer recharge, these results are an important initial finding regarding the estimation of recharge in the study area and can be used in water balance calculations for water management purposes. With further research into the observed relationships and parameterisation of these relationships, the soil moisture model could be updated to better represent recharge within this, and similar, study areas.</p>