Peatlands cover approximately three percent of Earth’s land surface, sequestering roughly one-third of terrestrial carbon while releasing nearly ten percent of atmospheric methane on an annual basis. Shifts in global temperature can alter these landscapes, from storing carbon under cold and waterlogged conditions, to rapidly expelling carbon during warmer and drier periods. In this study, we use eddy covariance datasets from the AmeriFlux Network (Lawrence Berkeley National Laboratory) to explore statistically significant spatial and temporal relationships between atmospheric variables and carbon fluxes, across two northern peatland systems with distinct hydrologic regimes: the Bonanza Creek Thermokarst Bog in Alaska, USA, and the Marcell Bog Lake Fen in Minnesota, USA. We set up a generalized multi-site long short-term memory (LSTM) model, trained on atmospheric and flux data, from four sites in Alaska in an effort to estimate methane fluxes at a fifth site, Bonanza Creek, using only atmospheric variables as inputs. For Marcell Bog Lake, we trained a complementary LSTM using a sequential 80:20 train–test split for estimation and evaluation. Half-hourly data from 2010 to 2020 was used for statistical analysis and to train each model. On average, Spearman correlation coefficients (ρ) revealed a subtle inverse relationship for methane flux and atmospheric pressure in winter months (ρ=-0.21), with considerable variability observed in spring and summer. Stronger positive correlations were shown for methane flux and air temperature (ρ=0.46), which remained relatively consistent across the four seasons. This was also apparent for methane concentration and humidity (ρ=0.58), suggesting that more saturated air tended to be associated with elevated methane levels. While overall model performance reflects the inherent challenges of cross-site prediction, the LSTM results exhibit a strong temporal agreement with flux observations for Bonanza Creek and Marcell Bog Lake, respectively, demonstrating promising skill for flux estimates in areas without dedicated instrumentation. This study explores atmospheric controls on methane flux estimates across thermokarst and fen landscapes, separated by approximately 4500 km, to better understand carbon cycling in ungauged northern peatland complexes.

Extreme weather event prediction remains a critical challenge in operational forecasting, with conventional methods demonstrating limited accuracy for intensity change detection. This study presents complementary advances through integrated machine learning frameworks and experimental mathematical analysis techniques for enhanced environmental prediction capabilities. We developed computational systems by combining neural network architectures with multi-source meteorological datasets from comprehensive Atlantic Basin storm records spanning a decade, representing near-complete population coverage. Machine learning algorithms achieved substantial overall accuracy with improved detection rates for critical intensity changes, significantly exceeding historical operational capabilities. Neural network trajectory prediction demonstrated competitive performance with mean absolute errors comparable to current operational standards across intensity categories. Most significantly, experimental mathematical analysis achieved enhanced prediction accuracy, representing substantial improvements over conventional environmental parameter methods. Using advanced geometric techniques on multi-dimensional atmospheric state representations, we identified structural patterns with the majority of critical events preceded by characteristic transitions within operationally relevant timeframes. This approach provided novel precursor signals for dangerous intensity changes. Feature analysis revealed that dynamic atmospheric variables dominated prediction importance versus static measurements, challenging traditional approaches. Validation identified optimal behavioral classifications corresponding to distinct meteorological regimes. These advances offer immediate operational value for weather forecasting while establishing theoretical foundations for understanding atmospheric systems as dynamical processes with discrete organizational phases, providing enhanced prediction capabilities essential for protecting vulnerable communities from climate change.

Hydrology Tool Set (HTS; http://www.hydrotools.tech) integrates a suite of free online hydrology tools and models. These can be used for advancing the understanding of local and watershed scale hydrological processes and can be employed in many areas of environmental research, such as the assessment of the impact of agricultural practices, urbanization, climate change, etc. All HTS tools and models are free to use and do not require user registration. HTS currently includes SepHydro (surface runoff and groundwater contributions to streamflow via hydrograph separation), ETCalc (evapotranspiration forms), SWIB (crop water deficit and excess, irrigation requirement and its impact on aquifer storage), SNOSWAB (snowpack processes, soil water content, soil water balance), TotPrePart (precipitation partitioning) and GWRech (groundwater recharge). These tools and models have been developed through a collaborative research effort of Environment and Climate Change Canada (ECCC), Agriculture and Agri-Food Canada (AAFC), Canadian Rivers Institute (CRI) and the University of New Brunswick (UNB). All the tools and models operate using daily datasets; have a user-friendly interface, with streamlined and easy to follow procedures; require minimal input data and provide flexibility with respect to the choice of methods, settings, displaying and exporting of data via plots and tables. The tools use a daily time step input and can be applied to any location for which data is available. The HTS tools and models can be used for scientific purposes, either for independent analyses or for providing input to more complex models, as well as for educational purposes, since each application includes a user guide and a sample data set, which can be used for familiarizing with the tool operation and/or for providing basic knowledge regarding key hydrological cycle components and processes. As of July 2025, ~60,000 HTS analyses have been conducted by users from around the world.

California’s large urban populations and valuable agricultural sectors are increasingly vulnerable to intensifying droughts driven by climate variability and change. Droughts are typically assessed in a sector-specific manner, with meteorological drought indices applied to urban systems and soil moisture-based indices used for agricultural impacts. However, meteorological and agricultural droughts are interconnected in space and time through soil moisture dynamics, which are critical to understand for effective cross-sector water management and climate change adaptation. We quantify meteorological drought using the Standardized Precipitation Index (SPI) and agricultural drought using the Standardized Soil Moisture Index (SSMI). We evaluate Coupled Model Intercomparison Project Phase 6 (CMIP6) and Variable Infiltration Capacity (VIC) modeled historical (1950-2014) and projected (2015 - 2100) drought across California, with emphasis on spring and summer months, when drought impacts are most acute across both urban and agricultural systems. We examine compound drought as the co-occurrence of meteorological and agricultural drought, represented by the SPI and SSMI, respectively, to identify periods when both conditions overlap and intensify and to assess how the frequency, severity, and temporal alignment of their joint occurrence change over time. This work introduces a novel conceptualization of compound drought as joint susceptibility to soil moisture-mediated meteorological and agricultural droughts impacting urban and agricultural sectors with connected water supply sources and thus compound risk.

A document by Wei Ji Leong. Click on the document to view its contents.

The magnetopause represents the boundary layer between Earth's magnetic field and the solar wind and plays a crucial role in mediating the transport of the solar wind's energy into the Earth's magnetosphere through the process of magnetic reconnection. Magnetopause properties, magnetospheric plasma outflows, and mesoscale structures along the magnetopause surface all can dampen magnetic reconnection and limit this energy transfer. To resolve this process and the effects that these mesoscale phenomena have at the magnetopause, we propose the Compression and Reconnection Investigations of the Magnetopause (CRIMP) mission as a Helio MIDEX A/O targeted concept mission designed to study the interactions between the Earth's magnetosphere and the solar wind in the magnetopause boundary layer. CRIMP targets open goals from the Heliophysics Decadal Survey by uncovering how local mass density enhancements affect global magnetic reconnection, how mesoscale structures drive magnetopause dynamics, and if the magnetopause acts as a perfectly absorbing boundary for radiation belt electrons through innovative in situ, multi-point, contemporaneous measurements. CRIMP was designed by 18 early career researchers in collaboration with Team X at NASA JPL as a part of the NASA Heliophysics Mission Design School.Presented at the 2024 AGU Fall Meeting, Advancing Fundamental Space Weather Research: Flares, CMEs, Solar Wind, and Earth/Planetary Impacts II Poster Session 

Accurate quantification of uncertainty and error in flood prediction is crucial for informed decision-making process. This study investigates the application of various uncertainty quantification approaches for flood prediction using state-of-the-art neural network models. By incorporating techniques such as Bayesian inference, Monte Carlo approach, and quantile regression, we aim to provide a comprehensive assessment of predictive uncertainty for flood prediction across multiple timescales. Our results highlight the comparative effectiveness of these models under different flood magnitudes, emphasizing their strengths and limitations across scales. This research contributes to improving the reliability of flood prediction models, thereby enhancing decision-making processes for water resource management and disaster preparedness. We expect that the findings will offer valuable insights into the integration of advanced deep learning techniques with uncertainty quantification to reduce error in short-long-term flood predictions.

A fluvial system and its response to allogenic controlled factors, such as sea level changes, tectonism, and climate variations is crucial for sediment transport system in the context of source-sink and paleoclimate research. Continental weathering and sea level changes have significantly influenced the sediment transport system of the Indus delta over the last two glacial-interglacial cycles.

Volatility is a critical factor in agri-commodity derivative trading, especially as the contract maturity date approaches. This study explores how volatility affects future price variations in the Indian agri-commodity derivatives market. We analyzed daily futures prices, spot prices, trading volumes, and open interest data for Farmer’s Producer Organization FPO-traded agri-commodities from January 2016 to December 2023. Our methodology included the Jonckheere–Terpstra test and Ordinary Least Squares (OLS) regressions with realized volatility to assess the impact of contract maturity on commodity prices. We also applied the Generalized Autoregressive Conditional Heteroskedasticity (GARCH) (1, 1) model to examine how volatility changes with the time to maturity of the derivative contracts. Additionally, we tested the negative covariance hypothesis to explore whether futures volatility in agri-commodities increases as maturity decreases. Our analysis, controlling for seasonality and time to maturity, found limited support for this hypothesis in the Indian futures market, except for barley and turmeric. We noted a positive correlation between close price volatility and trading volume, indicating that the rate of information arrival significantly affects volatility. These findings underscore the impact of volatility, time to maturity, and contact premium on hedging strategies and highlight opportunities to enhance trade policies for effective agri-derivative trading, benefiting rural farmer producers and other market participants.

Identifying the climate-induced variability in the condition of vegetation is particularly important in the context of recent climate change and plants’ impact on the mitigation of climate change. Here we present the coherence and time lags in the spectral response of individual forest types in the European temperate zone to the influencing meteorological factors in the period 2002–2022. Vegetation condition in broadleaved and coniferous forest was measured with monthly anomalies of two spectral indices – normalised difference vegetation index (NDVI) and enhanced vegetation index (EVI). As meteorological elements we used monthly anomalies of temperature (T), precipitation (P), vapour pressure deficit (VPD), evapotranspiration (ETo), and the teleconnection indices North Atlantic Oscillation (NAO) and North Sea Caspian Pattern (NCP). Periodicity in the time series was assessed using the wavelet transform, but no significant intra- or interannual cycles were detected in both vegetation (NDVI/EVI) and meteorological variables. In turn, coherence between NDVI/EVI and meteorological elements was described using the methods of wavelet coherence and Pearson’s linear correlation with time lag. In the European temperate zone analysed in this study, NAO produces strong coherence in a circa 1-year band and a weaker coherence in a circa 3-year band. The strongest relationships occur between condition of the forest vegetation and T and ETo – there is a significant cohesion with the 8–16-month (ca. 1-year) and 20–32-month (ca. 2-year) bands. Aa the results obtained for NDVI and EVI are not entirely consistent, we additionally coupled NDVI and EVI with the data on the amount of dead trees removed within sanitation felling in the shorter period 2015–2022, in order to determine which satellite-borne index represents better the actual condition of vegetation in forests of the European temperate zone. NDVI is a good predictor of the forest vegetation, substantially better than EVI. Spatially, NDVI gives more pixels with negative slope of the trend in spectral index values, while EVI seems to overestimate the number of positive slopes. The bigger number of negative slopes in the trend in NDVI seems to be in agreement with increasing volume of dead trees in the analysed 2015–2022 period.

A significant fraction of the Earth’s mantle is bordered on the top by the continental lithosphere and on the bottom by the Large Low Velocity Provinces(LLVPs), both of which can affect the efficiency of convective heat transfer within the mantle. Previous studies find that continents act like insulators for the mantle, buffering the heat leaving the system and causing mantle temperatures to rise over time. In contrast, research focused on LLVPs finds reduced heat transport from the core to the mantle consequently lowering mantle temperatures over time. However, these studies have only considered mantle convection with either top or bottom insulators, not both working together. Thus, the combined impact of competing basal and surface insulators on the global thermal state and mantle dynamics remains unclear. Here, we undertake a geodynamic study to quantify the effects of a dual-insulated system on mantle convection and Earth’s thermal evolution. We present preliminary results from a series of mantle convection simulations run in the finite-element code ASPECT, incorporating both the lithosphere and LLVPs, along with various Earth-like complexities in rheological and thermo-chemical structure including 1) the role of strongly temperature- and pressure-dependent linear and non-linear rheology, 2) radiogenic heating with different concentrations of HPEs in the mantle, continent, and LLVP layers, 3) differencesin convective vigor, and 4) the effect of mid- and deep-mantle phase changes. We aim to showcase the effects of variable heating and rheology on the mantle temperature and convection dynamics in the presence of dual insulators. The computational results will be compared between each set of conditions and with accompanying research on variable lithosphere and LLVP thickness and surface coverage for the isoviscous and uniform heating case.

Field-aligned currents are the main physical mechanism behind magnetosphere–ionosphere coupling, with these currents communicating environmental changes between the two regions of geospace. Field-aligned currents are thus a key factor in monitoring space weather, both from an operational and scientific perspective. While detecting field-aligned currents from the ground is challenging, their magnetic perturbations are easily detected by spacecraft in low-Earth orbit. The AMPERE project demonstrated that engineering magnetometers on commercial satellite constellations (in their case the Iridium constellation) can be assimilated to estimate the field-aligned currents across the entire polar cap. A limiting factor in the spatiotemporal resolution of this assimilation is the coverage of measurement points – Iridium consists of 66 satellites across only 6 orbital planes. However, in recent years much larger satellite mega-constellations have been launched into low-Earth orbit. Here we leverage engineering magnetometer data from the OneWeb constellation – 636 satellites distributed across 12 orbital planes – for the first time. A case study during the 12 May 2021 geomagnetic storm is presented. We discuss instrumentation, data availability, magnetic cleanliness, and calibration steps. Following this pre-processing of the OneWeb data, we show evidence of field-aligned current signatures present during this event which agree well with AMPERE maps. Finally, we provide an outlook on potential scientific and operational prospects that mega-constellations such as OneWeb might enable.

The Atlantic Coastal Plain is comprised of thick, unconsolidated sediments that span much of the East Coast of the United States. Seismic waves passing through this shallow, loosely compacted layer of sedimentary rock display increased ground motion amplitudes as measured on broadband seismometers deployed during the Southeastern Suture of the Appalachian Margin Experiment (SESAME) from 2010–2014. This amplification can lead to greater damage during earthquakes, making site response analysis, and constraints on the associated shallow shear velocity structure, important for informing ground motion modeling in the region. Site amplification is most noticeable in the horizontal direction and, therefore, can be quantified at single seismic stations by dividing the horizontal spectrum by the vertical spectrum of ambient seismic noise. The resulting horizontal-to-vertical spectral ratio (HVSR) provides a simple estimate of the fundamental frequency of a given site; however, more holistic modeling of the complete HVSR curve to constrain shallow elastic structure remains challenging. We leverage the diffuse field assumption (DFA) to interpret HVSR in terms of the ratios of the imaginary parts of the Green’s function components, which allows for the inversion of HVSR curves for elastic structure. We measure HVSR at 91 broadband stations of the Southeastern Suture of the Appalachian Margin Experiment (SESAME) and apply the HV-Inv tool to model the Atlantic Coastal Plain sediment structure beneath each station. While we observe clear lateral variability in HVSR across the array due to variations in sediment structure, we generally find little azimuthal variability of HVSR at most stations, allowing local simplification of the structure to horizontal layers. To resolve tradeoffs in the elastic structure due to fitting HVSR alone, we plan to incorporate additional observations of high-frequency Rayleigh wave ellipticity and multimode phase velocity estimated from ambient noise cross-correlations. Together, these complementary datasets will contribute to a robust framework for subsurface characterization in complex geological settings.

The dynamic equilibrium of a river basin can be modified by both natural as well as anthropogenic factors. Brahmani River Basin, bounded by the Mahanadi Rift and Chotanagpur plateau in the Indian shield, is the one of the largest river systems in India. In this study, various geomorphic properties of the 280 sub-watersheds of the Brahmani River basin have been investigated using ArcGIS and Matlab. This study aims to quantify evolutionary traits by investigating detailed geomorphic indices. Aside from the indices, longitudinal profiles and the channel steepness index (ksn) and knickpoint analysis have been evaluated to examine the river’s response to lithologic control and their relationship to the underlying reactivated structural features. Additionally, disequilibrium and transience dynamics have been measured, which revealed that most sub-basins are tectonically enhanced, naturally extended, and structurally organized. Besides this, reconstruction of paleo-relief of rivers longitudinal profiles have been done. Further, evidence also supports that anthropogenic stresses have been responsible for sediment transport during the Anthropocene. Statistical analysis such as the Mann-Kendall test, Pettitt test, Sen’s Slope, and double mass curve of recent suspended sediment load (1980 to 2021) at the terminal station also helps to detect the sudden change point. Collectively, the findings highlight a landscape intricately modified by both natural forces and human activities, underscoring the importance of integrated management and conservation strategies in response to these influences. Keywords: Geomorphic Indices, Sediment load, Lithology, Topotoolbox, Peninsular India.

Climate change mitigation strategies increasingly focus on negative emission technologies, including enhanced weathering. This study investigates the relationship between cation release in soil and climate factors such as rainfall and temperature, as a method of modeling the dissolution of rock from a enhanced rock weathering deployment. We conducted a field experiment comparing basalt-amended plots with a control plot, measuring cation concentrations in soil porewater over time. Concurrently, we recorded rainfall and temperature data to assess environmental influences on cation release dynamics. Results show significant differences in cation release between basalt-amended and control plots. Statistical modeling reveals strong correlations between cation concentrations and both rainfall and temperature. Our findings demonstrate strong correlations with climatic factors which underscores the importance of considering environmental conditions when implementing enhanced weathering techniques. This research contributes to our understanding of the efficacy and dynamics of enhanced weathering in real-world conditions, informing future large-scale applications of this climate change mitigation approach. The predictive model developed from this study offers a valuable tool for estimating the potential effectiveness of enhanced weathering across diverse environmental settings.

A document by Tasfia Tasnim. Click on the document to view its contents.

The enhanced severity and frequency of extreme weather events, such as combined heat waves and heavy rainfall, pose significant concerns due to their devastating regional impacts. This research quantifies the association between heatwaves and heavy rainfall events using Event Coincidence Analysis during 1951-2020. We compute both the Precursor Coincidence Rate (PCR) and the Trigger Coincidence Rate (TCR), which are key metrics in understanding the physical link between heatwaves and subsequent heavy rainfall events. The PCR indicates the proportion of heatwave events that precede heavy rainfall events, providing insight into the reliability of the precursor (heatwave events) as an indicator of the trigger event (heavy rainfall events). TCR represents the proportion of heavy rainfall events that follow heatwave events, useful for evaluating the predictive power of the precursor. Our results reveal TCR decreased in the recent period as compared to the base period for shorter time frame (temporal windows ΔT=2 and Delta ΔT=5), however, exhibited a notable rise (50%) observed in approximately 70% of the grid points for longer temporal windows ( ΔT=7). The observation is very well in congruence that the moisture holding capacity of the atmosphere has enhanced and convective activity and latent heat release (heatwave producing extreme rainfall) take longer than usual. Moreover, PCR increased for all temporal windows, indicating an increased probability of heatwaves triggering precipitation events within 7 days of their end and predictive potential in the recent period compared to the base period. Approximately 27% of grid points experienced one heat-wave event within 7 days prior to heavy rainfall occurrence in the recent period, compared to 13% in the base period, for every 5 precipitation events. Results indicate an increased coupling between the occurrence of heat waves and heavy rainfall within a very short period, attributed to moisture convergence and atmospheric convection.

The Himalayas are globally recognized as a biodiversity hotspot, harbouring a rich array of species. However, this ecological wealth of the region is under significant threat from anthropogenic activities and climate change. Key challenges to Himalayan biodiversity include deforestation, habitat fragmentation, overgrazing, and unsustainable tourism practices. Conservation efforts in the Himalayas are multifaceted and involve collaboration among local communities, governments, and international organizations. A fundamental step in any conservation initiative is species mapping, which helps prioritize limited conservation resources by identifying areas of high conservation value. The Himalayan landscape presents heterogeneity within species due to variations in slope, elevation, and aspect, further complicating conservation efforts. This study employs a semi-hypertemporal (SH) approach to map Pinus roxburghii, Quercus leucotricophora, and Alnus nepalensis in the Himalayan landscape surrounding the Dudhatoli forest area of Uttarakhand. Sixteen SH Sentinel-2 images between January and December 2022 were utilized. The normalized difference vegetation index (NDVI) was computed for all images to reduce the dimensionality of the dataset, creating an SH NDVI database. This database was then processed using various classification algorithms, including Maximum Likelihood Classification (MLC), Support Vector Machine (SVM), and fuzzy machine learning (ML) modified possibilistic c-means (MPCM) models with both conventional mean and the novel individual sample as mean (ISM) approaches. While ML and SVM are well-established methods known for their robustness and accuracy, they often struggle with the complexity and heterogeneity of mountainous landscapes. Our results indicate that the fuzzy MPCM ISM training approach outperforms ML and SVM classifiers, as well as the conventional mean approach. This is reflected in the overall accuracy values: 51% with MLC, 65.5% with SVM, 66.7% with conventional MPCM, with 82% for MPCM ISM training approach. The ISM approach effectively addresses the heterogeneity within the same class in the landscape. This study highlights the superiority of the MPCM ISM method contributing to more precise species mapping and informed conservation strategies.