Data assimilation plays a dual role in advancing the “scientific” understanding and serving as an “engineering tool” for the Earth system sciences. Land data assimilation (LDA) has evolved into a distinct discipline within geophysics, facilitating the harmonization of theory and data and allowing land models and observations to complement and constrain each other. Over recent decades, substantial progress has been made in the theory, methodology, and application of LDA, necessitating a holistic and in‐depth exploration of its full spectrum. Here, we present a thorough review elucidating the theoretical and methodological developments in LDA and its distinctive features. This encompasses breakthroughs in addressing strong nonlinearities in land surface processes, exploring the potential of machine learning approaches in data assimilation, quantifying uncertainties arising from multiscale spatial correlation, and simultaneously estimating model states and parameters. LDA has proven successful in enhancing the understanding and prediction of various land surface processes (including soil moisture, snow, evapotranspiration, streamflow, groundwater, irrigation and land surface temperature), particularly within the realms of water and energy cycles. This review outlines the development of global, regional, and catchment‐scale LDA systems and software platforms, proposing grand challenges of generating land reanalysis and advancing coupled land‒atmosphere DA. We lastly highlight the opportunities to expand the applications of LDA from pure geophysical systems to coupled natural and human systems by ingesting a deluge of Earth observation and social sensing data. The paper synthesizes current LDA knowledge and provides a steppingstone for its future development, particularly in promoting dual driven theory‐data land processes studies.

Land surface temperature (LST) is an important parameter to evaluate environmental changes. In this paper, time series analysis was conducted to estimate the interannual variations in global LST from 2001 to 2016 based on moderate resolution imaging spectroradiometer (MODIS) LST, and normalized difference vegetation index (NDVI) products and fine particulate matter (PM_2.5) data from the Atmospheric Composition Analysis Group. The results showed that LST, seasonally integrated normalized difference vegetation index (SINDVI), and PM_2.5 increased by 0.17 K, 0.04, and 1.02 μg/m³ in the period of 2001–2016, respectively. During the past 16 years, LST showed an increasing trend in most areas, with two peaks of 1.58 K and 1.85 K at 72°N and 48°S, respectively. Marked warming also appeared in the Arctic. On the contrary, remarkable decrease in LST occurred in Antarctic. In most parts of the world, LST was affected by the variation in vegetation cover and air pollutant, which can be detected by the satellite. In the Northern Hemisphere, positive relations between SINDVI and LST were found; however, in the Southern Hemisphere, negative correlations were detected. The impact of PM_2.5 on LST was more complex. On the whole, LST increased with a small increase in PM_2.5 concentrations but decreased with a marked increase in PM_2.5. The study provides insights on the complex relationship between vegetation cover, air pollution, and land surface temperature.

In situ soil moisture sensors provide highly accurate but very local soil moisture measurements, while remotely sensed soil moisture is strongly affected by vegetation and surface roughness. In contrast, cosmic-ray neutron sensors (CRNSs) allow highly accurate soil moisture estimation on the field scale which could be valuable to improve land surface model predictions. In this study, the potential of a network of CRNSs installed in the 2354 km² Rur catchment (Germany) for estimating soil hydraulic parameters and improving soil moisture states was tested. Data measured by the CRNSs were assimilated with the local ensemble transform Kalman filter in the Community Land Model version 4.5. Data of four, eight and nine CRNSs were assimilated for the years 2011 and 2012 (with and without soil hydraulic parameter estimation), followed by a verification year 2013 without data assimilation. This was done using (i) a regional high-resolution soil map, (ii) the FAO soil map and (iii) an erroneous, biased soil map as input information for the simulations. For the regional soil map, soil moisture characterization was only improved in the assimilation period but not in the verification period. For the FAO soil map and the biased soil map, soil moisture predictions improved strongly to a root mean square error of 0.03 cm³ cm⁻³ for the assimilation period and 0.05 cm³ cm⁻³ for the evaluation period. Improvements were limited by the measurement error of CRNSs (0.03 cm³ cm⁻³). The positive results obtained with data assimilation of nine CRNSs were confirmed by the jackknife experiments with four and eight CRNSs used for assimilation. The results demonstrate that assimilated data of a CRNS network can improve the characterization of soil moisture content on the catchment scale by updating spatially distributed soil hydraulic parameters of a land surface model.

Land surface models (LSMs) use a large cohort of parameters and state variables to simulate the water and energy balance at the soil–atmosphere interface. Many of these model parameters cannot be measured directly in the field, and require calibration against measured fluxes of carbon dioxide, sensible and/or latent heat, and/or observations of the thermal and/or moisture state of the soil. Here, we evaluate the usefulness and applicability of four different data assimilation methods for joint parameter and state estimation of the Variable Infiltration Capacity Model (VIC-3L) and the Community Land Model (CLM) using a 5-month calibration (assimilation) period (March–July 2012) of areal-averaged SPADE soil moisture measurements at 5, 20, and 50 cm depths in the Rollesbroich experimental test site in the Eifel mountain range in western Germany. We used the EnKF with state augmentation or dual estimation, respectively, and the residual resampling PF with a simple, statistically deficient, or more sophisticated, MCMC-based parameter resampling method. The performance of the calibrated LSM models was investigated using SPADE water content measurements of a 5-month evaluation period (August–December 2012). As expected, all DA methods enhance the ability of the VIC and CLM models to describe spatiotemporal patterns of moisture storage within the vadose zone of the Rollesbroich site, particularly if the maximum baseflow velocity (VIC) or fractions of sand, clay, and organic matter of each layer (CLM) are estimated jointly with the model states of each soil layer. The differences between the soil moisture simulations of VIC-3L and CLM are much larger than the discrepancies among the four data assimilation methods. The EnKF with state augmentation or dual estimation yields the best performance of VIC-3L and CLM during the calibration and evaluation period, yet results are in close agreement with the PF using MCMC resampling. Overall, CLM demonstrated the best performance for the Rollesbroich site. The large systematic underestimation of water storage at 50 cm depth by VIC-3L during the first few months of the evaluation period questions, in part, the validity of its fixed water table depth at the bottom of the modeled soil domain.

Drought can cause severe agricultural economic losses and hinder social and economic development. To manage drought, the process of drought events needs to be described with the help of an effective drought indicator. As a comprehensive variable, soil moisture is an essential indicator for describing agricultural drought. In this work, the extreme drought events in southwest China were analysed by the Global Land Data Assimilation System (GLDAS) root zone soil moisture data set. To define the drought quantitatively, a Standardized Soil Moisture Drought Index (SSMI) was calculated using the soil moisture data, then used to get the duration, frequency, and severity of drought events in southwest China. The results showed that the frequency and intensity of drought in southwest China had an apparent upward trend before 2014 and an apparent downward trend since 2014. Moreover, there are apparent differences in the frequency and intensity of drought in various regions of southwest China. Yunnan Province is prone to spring drought events. Guangxi Province and Guizhou Province are prone to spring, autumn and winter droughts, and the intensity of autumn and winter droughts is significantly higher than that of spring droughts. The Sichuan-Chongqing border area is prone to summer drought. We found that the monthly variation of soil moisture in different provinces in southwest China is consistent, but the seasonal variation of drought is different. Meanwhile, the performance of the SSMI was compared to the commonly used drought indices, the Standardized Precipitation Evapotranspiration Index (SPEI) and the Palmer Drought Severity Index (PDSI). The results showed that the SSMI is more sensitive to drought than both SPEI and PDSI in southwest China. The results also demonstrate that GLDAS soil moisture data can be used to study drought at a small regional scale.

Possible environmental change and ecosystem degradation have received increasing attention since the construction of Three Gorges Reservoir Catchment (TGRC) in China. The advanced Google Earth Engine (GEE) cloud-based platform and the large number of Geosciences and Remote Sensing datasets archived in GEE were used to analyze the land use and land cover change (LULCC) and climate variation in TGRC. GlobeLand30 data were used to evaluate the spatial land dynamics from 2000 to 2010 and Landsat 8 Operational Land Imager (OLI) images were applied for land use in 2015. The interannual variations in the Land Surface Temperature (LST) and seasonally integrated normalized difference vegetation index (SINDVI) were estimated using Moderate Resolution Imaging Spectroradiometer (MODIS) products. The climate factors including air temperature, precipitation and evapotranspiration were investigated based on the data from the Global Land Data Assimilation System (GLDAS). The results indicated that from 2000 to 2015, the cultivated land and grassland decreased by 2.05% and 6.02%, while the forest, wetland, artificial surface, shrub land and waterbody increased by 3.64%, 0.94%, 0.87%, 1.17% and 1.45%, respectively. The SINDVI increased by 3.209 in the period of 2000-2015, while the LST decreased by 0.253 °C from 2001 to 2015. The LST showed an increasing trend primarily in urbanized area, with a decreasing trend mainly in forest area. In particular, Chongqing City had the highest LST during the research period. A marked decrease in SINDVI occurred primarily in urbanized areas. Good vegetation areas were primarily located in the eastern part of the TGRC, such as Wuxi County, Wushan County, and Xingshan County. During the 2000–2015 period, the air temperature, precipitation and evapotranspiration rose by 0.0678 °C/a, 1.0844 mm/a, and 0.4105 mm/a, respectively. The climate change in the TGRC was influenced by LULCC, but the effect was limited. What is more, the climate change was affected by regional climate change in Southwest China. Marked changes in land use have occurred in the TGRC, and they have resulted in changes in the LST and SINDVI. There was a significantly negative relationship between LST and SINDVI in most parts of the TGRC, especially in expanding urban areas and growing forest areas. Our study highlighted the importance of environmental protection, particularly proper management of land use, for sustainable development in the catchment.