Storage Estimates Research Articles

Two-step carbon storage estimation in urban human settlements using airborne LiDAR and Sentinel-2 data based on machine learning

The quantification of carbon storage (CS) within urban areas has become increasingly crucial for achieving global carbon neutrality. This study proposed a new approach to estimating CS in urban human settlements, building on an existing forestry biomass expansion factor (BEF)-based CS estimation approach, and tested it over Suwon, a city in South Korea. First, a tree canopy cover map was created using high resolution land cover data. The urban tree area ratio was then calculated to estimate the BEF-based CS (Step 1 CS), considering parameters from the national forest inventories. Since urban trees have different growing environments and characteristics than forests, the CS for human settlements (Step 2 CS) was estimated using Step 1 CS as well as structural and spectral information of trees from light detection and ranging (LiDAR) and Sentinel-2 data via machine learning (ML) techniques. The dependent variable of the ML models was the difference between the Step 1 CS and the reference CS, which was calculated using field-measured data and existing allometric equations for urban areas. Using both LiDAR and Sentinel-2 information, the random forest model outperformed the other ML models tested, with an R2 of 0.884, a root mean squared error of 0.432 tC/900 m2, and a lower sampling strategy sensitivity. Feature analysis revealed that incorporating structural and spectral information resulted in a more reliable CS estimation by considering different tree environmental characteristics for each grid cell in the model. This data-driven model utilizes remote sensing and ML and can generate spatially explicit CS maps of large urban areas. It can accurately quantify CS in human settlements and is expected to be easily applicable to other urban green spaces.

Estimation of soil water storage according to its multi-scale correlations with environmental factors

Soil water storage (SWS) is a crucial parameter for vegetation restoration and a variety of hydrological processes in semi-arid regions. Its dynamics reflect the integrated influence of different factors operating at various intensities and scales. In this study, we investigated the scale-specific correlations between SWS and six selected environmental factors along a 1340-m long sampling transect in a semi-arid catchment using multivariate empirical mode decomposition (MEMD). The six factors were bulk density, elevation, clay, silt, sand, and soil organic carbon content. The multivariate data series of the SWS and factors of interest were separated into seven intrinsic mode functions (IMFs) and residue. IMF1, IMF3, and IMF5 were identified as the primary scales according to the percentage of total variance (about 50%) in SWS under different soil moisture conditions. The scale-specific correlations between SWS and the corresponding factors varied with scale but were irrelevant to soil moisture condition. SWS at each IMF and residue could be estimated using the associated scale-specific environmental factors at the same IMF or residue. SWS at the measurement scale was satisfactorily estimated by summarizing all the predicted IMFs and residue, which outperformed measurement-scale SWS estimation based on multivariate stepwise linear regression between SWS and the associated factors. Soil moisture condition could affect the relative importance of the environmental factors in overall SWS estimation. In general, soil properties were the dominant predictor of SWS under different soil moisture conditions. MEMD provides a useful opportunity to deal with scale-specific correlations between various variables.

Mating roles, copulation duration and reproductive output in the hermaphroditic freshwater snail Planorbella trivolvis

ABSTRACT Hermaphrodites exhibit a wide variety of reproductive strategies, yet empirical studies lag behind theories about their evolution. Planorbella trivolvis, a simultaneous hermaphrodite capable of reciprocal copulation and rare self-fertilization, provides a unique model system for studying reproduction. This study assessed sex roles during mating, demonstrating that mating is primarily reciprocal, and investigated the relationship between physical coupling and reproductive output for single copulation events. Snails were isolated before sexual maturity and placed in 10-l aquaria in size-matched groups. Upon copulation, pairs were removed to small cups until they separated; they were then kept isolated and eggs counted until egg production ceased. The frequency of reciprocal copulation (both members of a pair produced egg masses following copulation) was 95% (35/37 productive matings), higher than previously reported. The duration of egg production ranged from 19 to 197 days (mean ± SE = 110.13 ± 6.3 days) with a sharp drop in production of egg masses and eggs per mass occurring between 15 and 20 weeks postmating, consistent with previous estimates of sperm storage. There was no relationship between mating duration and total egg production (a biomarker of sperm availability) or the duration of egg production (a marker of sperm longevity). Further investigation is needed to directly estimate sperm transfer and tease apart the duration of reciprocal physical contact from gamete and accessory protein transfer.

Region‐specific drivers cause low organic carbon stocks and sequestration rates in the saltmarsh soils of southern Scandinavia

AbstractSaltmarshes are known for their ability to act as effective sinks of organic carbon (OC) and their protection and restoration could potentially slow down the pace of global warming. However, regional estimates of saltmarsh OC storage are often missing, including for the Nordic region. To address this knowledge gap, we assessed OC storage and accumulation rates in 17 saltmarshes distributed along the Danish coasts and investigated the main drivers of soil OC storage. Danish saltmarshes store a median of 10 kg OC m−2 (interquartile range, IQR: 13.5–7.6) in the top meter and sequester 31.5 g OC m−2 yr−1 (IQR: 41.6–15.7). In a global context, these values are comparatively low. Soils with abundant clay (> 20%), older and stable saltmarshes in mesohaline settings, and with low proportion of algal organic material showed higher OC densities, stocks, and accumulation rates. Grazing led to significantly higher OC stocks than neighboring ungrazed locations, likely due to trampling modifying soil abiotic conditions (higher erosion‐resistance and higher clay content) that slow carbon decay. Scaling up, Danish saltmarsh soils, comprising about 1% of the country's area, have the potential to yearly capture up to 0.1% of Denmark's annual consumption‐based CO2 emissions. Our research expands the baseline data needed to advance blue carbon research and management in the Nordic region while highlighting the need for a more comprehensive approach to saltmarsh management that considers the full range of services of these ecosystems and does not only focus on climate benefits.

Estimation for Refined Carbon Storage of Urban Green Space and Minimum Spatial Mapping Scale in a Plain City of China

Current cities are not concrete jungles and deserts with sparse vegetation. Urban green space (UGS) appears widely in human activity areas and plays an important role in improving the human living environment and accumulates carbon storage. However, given the scattered distribution of UGS, studies on both the refined spatial estimation of carbon storage and appropriate mapping scale are still lacking. Taking the downtown area of Kaifeng, China, as the study area, this study verified the i-Tree Eco model on the basis of a field survey and accurately estimated the spatial carbon storage of UGS by combining it with remote sensing data, and finally, we obtained the minimum spatial mapping scale of UGS carbon storage by scaling. The results showed that (1) the total area of UGS in study area was 26.41 km2, of which the proportion of total area of residential area and park green spaces was about 50%. The area of UGS per capita in the study area is 40.49 m2. (2) Within the 123 survey samples, the proportion of communities with tree–shrub–herbs structure was the highest, 51.22%. The average carbon density was 5.89 kg m−2, among which the park, protective and square green spaces had the highest carbon density in all land use types. (3) The total carbon storage of UGS in the study area was 114,389.17 t, and the carbon storage of UGS per capita was 175.39 kg. Furthermore, the scaling analysis showed that 0.25 km spatial resolution was the minimum spatial scale for UGS carbon storage mapping. This study improves our understanding of urban carbon storage, highlights the role and potential of UGS in carbon neutrality, and clarifies the importance of estimating urban carbon storage at appropriate scales. This study is also of great significance for rationally understanding the terrestrial carbon cycle in urban areas and improving regional climate simulations.

Global groundwater droughts are more severe than they appear in hydrological models: An investigation through a Bayesian merging of GRACE and GRACE-FO data with a water balance model

Realistic representation of hydrological drought events is increasingly important in world facing decreased freshwater availability. Index-based drought monitoring systems are often adopted to represent the evolution and distribution of hydrological droughts, which mainly rely on hydrological model simulations to compute these indices. Recent studies, however, indicate that model derived water storage estimates might have difficulties in adequately representing reality. Here, a novel Markov Chain Monte Carlo - Data Assimilation (MCMC-DA) approach is implemented to merge global Terrestrial Water Storage (TWS) changes from the Gravity Recovery And Climate Experiment (GRACE) and its Follow On mission (GRACE-FO) with the water storage estimations derived from the W3RA water balance model. The modified MCMC-DA derived summation of deep-rooted soil and groundwater storage estimates is then used to compute 0.5∘ standardized groundwater drought indices globally to show the impact of GRACE/GRACE-FO DA on a global index-based hydrological drought monitoring system. Our numerical assessment covers the period of 2003–2021, and shows that integrating GRACE/GRACE-FO data modifies the seasonality and inter-annual trends of water storage estimations. Considerable increases in the length and severity of extreme droughts are found in basins that exhibited multi-year water storage fluctuations and those affected by climate teleconnections.

Shifts in internal stem damage along a tropical precipitation gradient and implications for forest biomass estimation.

Woody biomass is a large carbon store in terrestrial ecosystems. In calculating biomass, tree stems are assumed to be solid structures. However, decomposer agents such as microbes and insects target stem heartwood, causing internal wood decay which is poorly quantified. We investigated internal stem damage across five sites in tropical Australia along a precipitation gradient. We estimated the amount of internal aboveground biomass damaged in living trees and measured four potential stem damage predictors: wood density, stem diameter, annual precipitation, and termite pressure (measured as termite damage in downed deadwood). Stem damage increased with increasing diameter, wood density, and termite pressure and decreased with increasing precipitation. High wood density stems sustained less damage in wet sites and more damage in dry sites, likely a result of shifting decomposer communities and their differing responses to changes in tree species and wood traits across sites. Incorporating stem damage reduced aboveground biomass estimates by > 30% in Australian savannas, compared to only 3% in rainforests. Accurate estimates of carbon storage across woody plant communities are critical for understanding the global carbon budget. Future biomass estimates should consider stem damage in concert with the effects of changes in decomposer communities and abiotic conditions.

Quantifying Intermittent Flow Regimes in Ungauged Basins: Optimization of Remote Sensing Techniques for Ephemeral Channels Using a Flexible Statistical Classification

Intermittent and ephemeral channels are a critical component of the global hydrologic network. The dominant feature in dryland environments, ephemeral channel transmission loss facilitates aquifer recharge. Characterizing flow intermittency improves groundwater storage estimates; however, limited gauging of intermittent systems impedes this understanding. This research develops an improved classification for surface flow, optimized for ephemeral systems using linear discriminant function analysis and remotely sensed imagery. It further applies this methodology to assess temporal and spatial flow patterns across the Souss channel, an ungauged, ephemeral system in central Morocco. Linear discriminant function analysis demonstrates high predictive accuracy for Landsat imagery, with significantly improved classification success as compared to the Modified Normalized Difference Water Index. Application to the Souss channel from 1984 to 2022 points to a decreasing trend in flow frequency. Despite this change, flow events remain concentrated within the wet season, critical for regional aquifer recharge. Spatial flow characteristics further support sustained infiltration, with the majority of events focused within the upstream channel section during both dry and wet seasons. Decreased occurrence moving downstream highlights the likely impact of additional factors such as transmission loss, evapotranspiration, and agricultural abstraction contributing to channel intermittency.

Estimation of carbon storage of tropical mangroves in the North Maluku Islands region

The mangroves possess a remarkable capacity to absorb carbon dioxide (CO2), surpassing tropical rainforests by fourfold. This underscores the crucial role played by mangroves in global efforts to combat climate change. This study aimed to assess the biomass, carbon content, and CO2 absorption capabilities of mangrove ecosystems. Data collection followed the line transect method, utilizing the Mangrove Watch guide, which involved recording the diameter of mangroves at breast height (DBH), typically around 1.3 meters above the ground, as well as identifying the mangrove species and quantifying the number of stands for each tree species. The analysis of mangrove biomass, carbon content, and CO2 absorption capacity was conducted through an allometric modeling approach. The findings revealed the presence of seven distinct mangrove tree species at the study site, namely Rhizophora mucronata, R. apiculata, R. stylosa, Sonneratia alba, Bruguiera gymnorrhiza, Xylocarpus granatum, and Ceriops decandra. The mangrove density in this area was recorded at 1406 trees per hectare, with a corresponding biomass of 630,508 tons per hectare, a carbon content of 246,035 tons of carbon per hectare, and an impressive CO2 absorption capacity of 902,127 tons of CO2 per hectare. Notably, the mangrove vegetation in this research site exhibited a substantial capacity for CO2 absorption and biomass storage, surpassing many other locations in different regions. These findings underscore the significant role that island mangrove ecosystems can play in mitigating the effects of climate change.

Ecosystem-specific patterns and drivers of global reactive iron mineral-associated organic carbon

Abstract. Reactive iron (Fe) oxides are vital for long-term soil or sediment organic carbon (SOC) storage. However, the patterns and drivers of Fe-associated organic carbon (Fe-OC) over global geographic scales under various ecosystem types remain controversial. Here, we provided a systematic assessment of the distribution patterns and determinants of Fe-OC content and its contribution to SOC (fFe-OC) by assembling a global dataset comprising 862 observations from 325 sites in distinct ecosystems. We found that Fe-OC content across global ecosystems ranged from 0 to 83.3 g kg−1 (fFe-OC ranged from 0 % to 82.4 %), reflecting the high variability of the Fe-OC pool. Fe-OC contents varied with ecosystem type being greater in wetlands with a high molar ratio of Fe-OC / dithionite-extractable Fe (Fed) compared with marine and terrestrial ecosystems. Furthermore, fFe-OC in wetlands was significantly lower than that in other ecosystems due to rich organic carbon (OC). In contrast with climate variables and soil pH, the random forest modeling and multivariate analysis showed that the Fe-OC : Fed and SOC were the predominant predictors of Fe-OC content and fFe-OC in wetlands and terrestrial ecosystems, whereas Fed content was a primary driver in marine ecosystems. Based on upper estimates of global SOC storage in various ecosystem types, we further estimated that 83.84 ± 3.8, 172.45 ± 8.74, and 24.48 ± 0.87 Pg of SOC were preserved by association with Fe oxides in wetland, terrestrial, and marine ecosystems, respectively. Taken together, our findings highlighted the importance of reactive Fe oxides in global SOC preservation, and their controlling factors were ecosystem specific.

Research on Sustainable Shallow Geothermal Potential in Jinan Start-Up Area, China

Shallow geothermal energy (SGE), as an important renewable energy, playing an important role in reducing carbon emissions. In order to efficiently and sustainably utilize SGE, field investigation and storage estimation are needed. In this study, the hydrogeological data obtained from the field exploration of Jinan Start-up Area were collected and compiled. By analyzing the geotechnical property data and thermal response test results, the information of geotechnical and thermal properties and underground temperature distribution characteristics were collected. Subsequently, the analytic hierarchy process (AHP) combined with the comprehensive index method (CIM) were used to classify the shallow geothermal potential of Jinan Start-up Area. The entire area was divided into a high-potential area, medium-potential area and general area, of which 92.2% was high-potential area. The preliminary results, combined with the parameters obtained from the testing, indicate that the SGE storage at a borehole depth of 120 m is estimated to be approximately 2.68 × 1012 kJ·K−1, while the heat exchanger power of the buried pipe at the same depth is calculated to be around 1.73 × 105 kW. Finally, suggestions are given for sustainable development and utilization of SGE in this area.

Time-varying information measures: an adaptive estimation of information storage with application to brain-heart interactions.

Network Physiology is a rapidly growing field of study that aims to understand how physiological systems interact to maintain health. Within the information theory framework the information storage (IS) allows to measure the regularity and predictability of a dynamic process under stationarity assumption. However, this assumption does not allow to track over time the transient pathways occurring in the dynamical activity of a physiological system. To address this limitation, we propose a time-varying approach based on the recursive least squares algorithm (RLS) for estimating IS at each time instant, in non-stationary conditions. We tested this approach in simulated time-varying dynamics and in the analysis of electroencephalographic (EEG) signals recorded from healthy volunteers and timed with the heartbeat to investigate brain-heart interactions. In simulations, we show that the proposed approach allows to track both abrupt and slow changes in the information stored in a physiological system. These changes are reflected in its evolution and variability over time. The analysis of brain-heart interactions reveals marked differences across the cardiac cycle phases of the variability of the time-varying IS. On the other hand, the average IS values exhibit a weak modulation over parieto-occiptal areas of the scalp. Our study highlights the importance of developing more advanced methods for measuring IS that account for non-stationarity in physiological systems. The proposed time-varying approach based on RLS represents a useful tool for identifying spatio-temporal dynamics within the neurocardiac system and can contribute to the understanding of brain-heart interactions.

Streamflow Composition and Water “Imbalance” in the Northern Himalayas

AbstractThe Yarlung Zangbo River (YZR) is the largest river in the northern Himalayas, providing crucial water resources for downstream. A full understanding of the streamflow dynamics and regional water budget is critical to secure water security of the Himalayan water tower. Here we establish a comprehensive hydrological model to simulate the precipitation‐runoff‐evapotranspiration‐groundwater‐streamflow complex in the YZR basin. We decipher contributions of different water sources (e.g., precipitation, meltwater, groundwater) to YZR's streamflow and estimate that groundwater sustains ∼36% of annual streamflow in the YZR, while precipitation and melt surface runoff contribute 40% and 24%, respectively. Combining modeling, observation and reanalysis data, our results reveal a water “imbalance” that ∼31% of annual precipitation and meltwater (∼333 mm yr−1 or ∼85 km3 yr−1) is unaccounted for in the YZR basin. We propose that the “excess water” discharges to deep fractured bedrock aquifers, which is promoted by widespread permeable active structures (e.g., faults, fractures). This hypothesis is supported by groundwater storage (GWS) estimates where inclusion of the deep groundwater bridges the discrepancy between baseflow‐derived (shallow) GWS and those derived from the Gravity Recovery and Climate Experiment satellite data. The deep groundwater most likely flows across basins, bypasses streams, and finally discharges to downstream aquifers in the Indo‐Gangetic Plain as mountain block recharge. This study not only provides a comprehensive analysis of the streamflow composition in the YZR, but also contributes to shaping a more complete picture of the functionality of the Himalayan water tower, highlighting the importance of groundwater in regional water transfers.

Mesoscale soil moisture survey by mobile cosmic-ray neutron sensor across various landscapes in the Heihe River Basin

The cosmic-ray neutron technology is a novel method for upscaling soil moisture, holding significant importance in drought monitoring, soil water storage estimation, calibration of remote sensing products, and data assimilation of land surface models. However, most studies conducted soil moisture measurement experiments using static Cosmic-Ray Neutron Sensors (CRNS) and are often limited to a single landscape, lacking in-situ measurement for the cosmic-ray neutron rover across various landscapes. This study carried out soil moisture observation experiments using a cosmic-ray neutron rover in nine grassland plots, nine farmland plots, and nine desert plots of the Heihe River Basin, Northwest China. The neutron counts displayed clear variations among different plots, and an equation could represent the relationship between neutron intensity and multiple variables near the ground. The parameter N0 was the smallest for grassland and greatest for desert, and there was a significant negative correlation between Normalized Difference Vegetation Index and N0 (P <0.05). Furthermore, the performance of CRNS notably improved after considering soil lattice water and soil organic matter. The Root Mean Square Error (RMSE) between the measured and oven-dried soil moisture increased with radius. Soil moisture measured by CRNS strongly supported the 100 m diameter footprint assumption (RMSE <0.044 g/g). The study offers valuable insights into the application and promotion of cosmic-ray neutron technology across various landscapes within a basin.

Polyurethane shape memory polymer: structure characterization and estimation of energy storage and dissipation during the tension process

Polyurethane shape memory polymer: structure characterization and estimation of energy storage and dissipation during the tension process

MONITORING GROUNDWATER STORAGE BASINS AND HYDROLOGICAL CHANGES USING THE GRACE SATELLITE AND SENTINEL-1 FOR THE GANGA RIVER BASIN

Abstract. Groundwater depletion-related subsidence is a significant issue in many parts of the world. It can permanently reduce the amount of groundwater stored in an aquifer and even cause structural damage to the Earth’s surface. The Ganga Basin in the northwestern region of India is no exception, with around a meter of subsidence occurring between 2018 and 2023. However, understanding the connection between variations in groundwater quantities and ground deformation has been challenging. We used surface displacement measurements from InSAR and gravimetric terrestrial water storage estimates from the GRACE satellite pair to characterize the hydrological dynamics within the Ganga Basin. Sentinel-1 was used to map the entire Ganga River basin in the inundated zone. The InSAR time series shows coherent short-term changes that coincide with hydrological features when the long-term aquifer compaction is removed. For instance, an uplift is seen at the confluence of multiple rivers and streams that drain into the southeastern margin of the basin in the winters of 2018–2019 and 2021–2022. Imaging the monthly spatial variations in water volumes is based on these data and calculations of mass changes from the orbiting of Sentinel-1 and GRACE satellites. We even employ machine learning techniques as evaluative methods to make it simple to combine InSAR quickly and convincingly with gravimetric datasets, which will help advance global efforts to understand better and manage groundwater resources.

Defining Renewable Groundwater Use and Its Relevance to Sustainable Groundwater Management

AbstractGroundwater systems are commonly defined as renewable or non‐renewable based on natural fluxes of recharge or on estimates of aquifer storage and groundwater residence time. However, we show here that the principle of capture (i.e., how recharge and discharge change due to pumping) challenges simple definitions so that a groundwater system cannot be renewable or non‐renewable in and of itself, but only with reference to how the groundwater is being used. We develop and propose more hydraulically informed definitions for flux‐renewable and storage‐renewable groundwater use, and a combined definition that encompasses both the flux‐based and storage‐based perspectives such that: renewable groundwater use allows for dynamically stable re‐equilibrium of groundwater levels and quality on human timescales. Further, we show how a matrix of combinations of (a) the ratio of pumping rate to the maximum rate of capture along with (b) the response or recovery timescales implicit in this definition, leads to a useful four‐quadrant framework for characterizing groundwater use, illustrated using case studies from aquifers around the world. Renewable groundwater use may inform pathways to groundwater sustainability, which encompasses a broader set of dimensions (e.g., socio‐political, economic, ecological and cultural) beyond the scope of groundwater science. We propose that separating physically robust definitions of renewable groundwater use from the inherently value‐based language of sustainability, can help bring much needed clarity to wider discussions about sustainable groundwater management strategies, and the role of groundwater science and scientists in such endeavors.

Making the Best Use of GRACE, GRACE‐FO and SMAP Data Through a Constrained Bayesian Data‐Model Integration

AbstractThe Gravity Recovery and Climate Experiment (GRACE, 2003–2017) and its Follow‐On mission GRACE‐FO (2018‐now) provide global estimates of the vertically integrated Terrestrial Water Storage Changes (TWSC). Since 2015, the Soil Moisture Active Passive (SMAP) radiometer observes global L‐band brightness temperatures, which are sensitive to near‐surface soil moisture. In this study, we introduce our newly developed Constrained Bayesian (ConBay) optimization approach to merge the TWSC of GRACE/GRACE‐FO along with SMAP soil moisture data into the ∼10 km resolution W3RA water balance model. ConBay is formulated based on two hierarchical multivariate state‐space models to (I) separate land hydrology compartments from GRACE/GRACE‐FO TWSC, and (II) constrain the estimation of surface soil water storage based on the SMAP data. The numerical implementation is demonstrated over the High Plain (HP) aquifer in the United States between 2015 and 2021. The implementation of ConBay is compared with an unconstrained Bayesian formulation, and our validations are performed against in‐situ USGS groundwater level observations and the European Space Agency (ESA)'s Climate Change Initiative (CCI) soil moisture data. Our results indicate that the single GRACE/GRACE‐FO assimilation improves particularly the groundwater compartment. Adding SMAP data to the ConBay approach controls the updates assigned to the surface storage compartments. For example, correlation coefficients between the ESA CCI and the ConBay‐derived surface soil water storage (0.8) that are considerably higher than those derived from the unconstrained experiment (−0.3) in the North HP. The percentage of updates introduced to the W3RA groundwater storage is also decreased from 64% to 57%.

Large‐Scale Groundwater Monitoring in Brazil Assisted With Satellite‐Based Artificial Intelligence Techniques

AbstractHere, we develop and test an artificial intelligence (AI)‐based approach to monitor major Brazilian aquifers. The approach combines Gravity Recovery and Climate Experiment (GRACE) data and ground‐based hydrogeological measurements from Brazil's Integrated Groundwater Monitoring Network at hundreds of wells distributed in 12 aquifers across the country. We tested model ensembles based on three AI approaches: Extreme Gradient Boost, Light Gradient Boosting Model and CatBoost, followed by a Linear Regression step. The approach is further boosted with wavelet and seasonal decomposition processes applied to GRACE data. To determine the AI‐based model's sensitivity to data availability, we propose four experiments combining hydrogeological measurements from different aquifers. Groundwater storage (GWS) estimates from the Global Land Data Assimilation System (GLDAS) are used as benchmark. A sensitivity analysis shows that the LR‐based model ensemble is the best suited and to reproduce GWS change in all studied Brazilian aquifers. Results show that the proposed approach outperforms GLDAS in all experiments, with an root mean square error (RMSE) value of 2.68 cm for the experiment that covers all monitored wells in Brazil. GLDAS resulted in RMSE = 6.76 cm. Using our AI model outputs, we quantified the GWS change of two major aquifers, Urucuia and Bauru‐Caiuá, over the past two decades: −31 and −6 km3, respectively. Water loss is driven by a prolonged drought across most of the country and intensification of groundwater pumping for irrigation. This study demonstrates that combining satellite data and AI can be a cost‐effective alternative to monitor poorly equipped aquifers at the continental scale, with possible global replicability.

Frontiers of thermobarometry: GAIA, a novel Deep Learning-based tool for volcano plumbing systems

The anatomy of the plumbing system of active volcanoes is fundamental to understand how magma is stored and channeled to the surface. Reliable geothermobarometric estimates are, therefore, critical to assess the depths and temperatures of the complex system of magmatic reservoirs that form a volcano apparatus. Here, we developed a novel Machine Learning approach (named GAIA, Geo Artificial Intelligence thermobArometry) based upon Feedforward Neural Networks to estimate P-T conditions of magma (clinopyroxene) storage and migration within the crust. Our Feedforward Neural Network method applied to clinopyroxene compositions yields better uncertainties (Root-Mean-Square Error and R2 score) than previous Machine Learning methods and set the basis for a novel generation of reliable geothermobarometers, which extends beyond the paradigm associated to crystal-liquid equilibrium. Also, the bootstrap procedure, inherent to the Feedforward Neural Network architecture, permits to perform a rigorous assessment of the P-T uncertainty associated to each clinopyroxene composition, as opposed to the Root-Mean-Square Error representing the P-T uncertainty of whole set of clinopyroxene compositions.As a test, we applied GAIA to assess P-T conditions of five Italian volcanoes (Somma-Vesuvius, Campi Flegrei, Etna, Stromboli, Volcano), which are among the most dangerous volcanic centres in Europe. The results on the depths of the plumbing systems are in excellent agreement with those obtained with independent geophysical and geodetic surveys, and provide further evidence to current models of volcano plumbing systems consisting of physically-separated reservoirs interconnected by a network of conduits channelling magma en route to the surface. The results on the magma (clinopyroxene crystallization) temperatures are also in agreement with other estimates, albeit obtained considering - mainly but not only - thermodynamically-based clinopyroxene-liquid geothermometers.GAIA can set robust estimates of magma storage, segregation, and ascent conditions within the plumbing system of active volcanoes, helping to unravel P-T variations, if any, during their eruptive history and providing robust clues to volcanic hazard assessment.