Alfvén intermittency in space plasmas

The onset of Alfvén intermittent chaos in space plasmas is studied by numerically solving the derivative non-linear Schrödinger equation (DNLS) under the assumption of stationary Alfvén waves. We describe how the Alfvénic fluctuations of the magnetic field can evolve from periodic to chaotic behavior through a sequence of bifurcations as the plasma dissipation is varied. The collision of a chaotic attractor with an unstable periodic orbit leads to the generation of strongly chaotic behavior, in an event known as interior crisis. We also show that in the DNLS equation, chaotic attractors coexist with nonattracting chaotic sets responsible for transient chaotic behaviors. After the interior crisis point, a wide chaotic attractor can be decomposed into two coupled nonattracting chaotic sets, resulting in intermittent chaotic time series. Understanding transient chaos is a key to understand intermittency in space plasmas.     

Empirical evidences of persistence and dynamical chaos in solar–terrestrial phenomena

This paper presents a brief review of selected publications concerning dynamical chaos and persistence in various solar–terrestrial phenomena ranging from solar activity to climate dynamics. It draws attention to the advanced approaches known in many research areas (meteorology, hydrology, biology, economics, etc.), but not yet sufficiently used in solar–terrestrial physics. First, we introduce the concepts of dynamical (deterministic) chaos and fractional Brownian motion. Next, we discuss appropriate methods—fluctuation analysis and nonlinear time series analysis—for treatment of erratic time series based on these concepts. We outline some pitfalls and problems in the application of the discussed methods to empirical data. Finally, we present selected empirical evidence for persistence and dynamical chaos in solar activity, solar wind, magnetosphere and ionosphere, weather and climate systems.     

Soil–pile–structure interaction under SH wave excitation

A continuum model for the interaction analysis of a fully coupled soil–pile–structure system under seismic excitation is presented in this paper. Only horizontal shaking induced by harmonic SH waves is considered so that the soil–pile–structure system is under anti‐plane deformation. The soil mass, pile and superstructure were all considered as elastic with hysteretic damping, while geometrically both pile and structures were simplified as a beam model. Buildings of various heights in Hong Kong designed to resist wind load were analysed using the present model. It was discovered that the acceleration of the piled‐structures at ground level can, in general, be larger than that of a free‐field shaking of the soil site, depending on the excitation frequency. For typical piled‐structures in Hong Kong, the amplification factor of shaking at the ground level does not show simple trends with the number of storeys of the superstructure, the thickness and the stiffness of soil, and the stiffness of the superstructure if number of storeys is fixed. The effect of pile stiffness on the amplification factor of shaking is, however, insignificant. Thus, simply increasing the pile size or the superstructure stiffness does not necessarily improve the seismic resistance of the soil–pile–structure system; on the contrary, it may lead to excessive amplification of shaking for the whole system. Copyright © 2003 John Wiley & Sons, Ltd.     

Comparison of P‐wave attenuation models of wave‐induced flow

Wave‐induced oscillatory fluid flow in the vicinity of inclusions embedded in porous rocks is one of the main causes for P‐wave dispersion and attenuation at seismic frequencies. Hence, the P‐wave velocity depends on wave frequency, porosity, saturation, and other rock parameters. Several analytical models quantify this wave‐induced flow attenuation and result in characteristic velocity–saturation relations. Here, we compare some of these models by analyzing their low‐ and high‐frequency asymptotic behaviours and by applying them to measured velocity–saturation relations. Specifically, the Biot–Rayleigh model considering spherical inclusions embedded in an isotropic rock matrix is compared with White's and Johnson's models of patchy saturation. The modeling of laboratory data for tight sandstone and limestone indicates that, by selecting appropriate inclusion size, the Biot‐Rayleigh predictions are close to the measured values, particularly for intermediate and high water saturations.     

Evaluating the dual‐boundary forcing concept in subsurface–land surface interactions of the hydrological cycle

Subsurface and land surface processes (e.g. groundwater flow, evapotranspiration) of the hydrological cycle are connected via complex feedback mechanisms, which are difficult to analyze and quantify. In this study, the dual‐boundary forcing concept that reveals space–time coherence between groundwater dynamics and land surface processes is evaluated. The underlying hypothesis is that a simplified representation of groundwater dynamics may alter the variability of land surface processes, which may eventually affect the prognostic capability of a numerical model. A coupled subsurface–land surface model ParFlow.CLM is applied over the Rur catchment, Germany, and the mass and energy fluxes of the coupled water and energy cycles are simulated over three consecutive years considering three different lower boundary conditions (dynamic, constant, and free‐drainage) based on groundwater dynamics to substantiate the aforementioned hypothesis. Continuous wavelet transform technique is applied to analyze scale‐dependent variability of the simulated mass and energy fluxes. The results show clear differences in temporal variability of latent heat flux simulated by the model configurations with different lower boundary conditions at monthly to multi‐month time scales (~32–91 days) especially under soil moisture limited conditions. The results also suggest that temporal variability of latent heat flux is affected at even smaller time scales (~1–3 days) if a simple gravity drainage lower boundary condition is considered in the coupled model. This study demonstrates the importance of a physically consistent representation of groundwater dynamics in a numerical model, which may be important to consider in local weather prediction models and water resources assessments, e.g. drought prediction. Copyright © 2015 John Wiley & Sons, Ltd.     

River–aquifer interactions in a semi‐arid environment stressed by groundwater abstraction

Rivers and aquifers are, in many cases, a connected resource and as such the interactions between them need to be understood and quantified for the resource to be managed appropriately. The objective of this paper is to advance the understanding of river–aquifer interactions processes in semi‐arid environments stressed by groundwater abstraction. This is performed using data from a specific catchment where records of precipitation, evapotranspiration, river flow, groundwater levels and groundwater abstraction are analysed using basic statistics, hydrograph analysis and a simple mathematical model to determine the processes causing the spatial and temporal changes in river–aquifer interactions. This combined approach provides a novel but simple methodology to analyse river–aquifer interactions, which can be applied to catchments worldwide. The analysis revealed that the groundwater levels have declined (~ 3 m) since the onset of groundwater abstraction. The decline is predominantly due to the abstraction rather than climatic changes (r = 0.84 for the relationship between groundwater abstraction and groundwater levels; r = 0.92 for the relationship between decline in groundwater levels and magnitude of seasonal drawdown). It is then demonstrated that, since the onset of abstraction, the river has changed from being gaining to losing during low‐flow periods, defined as periods with flow less than 0.5, 1.0 or 1.5 GL/day (1 GL/day = 1 × 10⁶ m³/day). If defined as < 1.0 GL/day, low‐flow periods constitute approximately 65% of the river flows; the periods where the river is losing at low‐flow conditions are thus significant. Importantly, there was a significant delay (> 10 years) between the onset of groundwater abstraction and the changeover from gaining to losing conditions. Finally, a relationship between the groundwater gradient towards the river and the river flow at low‐flow is demonstrated. The results have important implications for water management as well as water ecology and quality. Copyright © 2012 John Wiley & Sons, Ltd.     

Comparison of two modeling approaches for groundwater–surface water interactions

An assessment of interactions between groundwater and surface water was carried out by applying two different modeling approaches to a small‐scale study area in the municipality of Havelock, Quebec. The first approach involved a commonly used sequential procedure that consists in determining the daily recharge rate using a quasi 2D infiltration model (HELP), applied in the next step as an imposed flux to a 3D finite‐element groundwater flow model. The flow model was calibrated under steady‐state and transient conditions against measured water levels. The second approach was based on a recently developed physically based, 3D fully coupled groundwater–surface water flow model (CATHY) applied to the entire flow domain in an integrated manner. Implementation, calibration, and results of the simulations for both approaches are presented and discussed. For equal annual precipitation (1038 mm/y) and evapotranspiration (556 mm/y), the second approach computed a recharge rate of 233 mm/y (8.9% higher than the first approach) and a net upward flow from the fractured aquifer (the first approach predicted a net downward flow to the rock). The simulated annual discharge was similar for the two approaches (9.6% difference). Both approaches were found to be useful in understanding the interactions between groundwater and surface water, although limitations are apparent in the sequential procedure's inability to account for surface–subsurface feedbacks, for instance near stream reaches where groundwater discharge is prevalent. The decoupled, two‐model approach provides disaggregated surface, vadose, and aquifer flows, and a simple aperçu at the different components of total discharge. The fully coupled model accounts for continuous water exchanges between the land surface, subsurface, and stream channel in a more complex manner, and produces a better match against observed data. Copyright © 2012 John Wiley & Sons, Ltd.     

Heterogeneity in ground water–surface water interactions in the hyporheic zone of a salmonid spawning stream

Spatial and temporal variability in ground water–surface water interactions in the hyporheic zone of a salmonid spawning stream was investigated. Four locations in a 150‐m reach of the stream were studied using hydrometric and hydrochemical tracing techniques. A high degree of hydrological connectivity between the riparian hillslope and the stream channel was indicated at two locations, where hydrochemical changes and hydraulic gradients indicated that the hyporheic zone was dominated by upwelling ground water. The chemistry of ground water reflected relatively long residence times and reducing conditions with high levels of alkalinity and conductivity, low dissolved oxygen (DO) and nitrate. At the other locations, connectivity was less evident and, at most times, the hyporheic zone was dominated by downwelling stream water characterized by high DO, low alkalinity and conductivity. Substantial variability in hyporheic chemistry was evident at fine (<10 m) spatial scales and changed rapidly over the course of hydrological events. The nature of the hydrochemical response varied among locations depending on the strength of local ground water influence. It is suggested that greater emphasis on spatial and temporal heterogeneity in ground water–surface water interactions in the hyporheic zone is necessary for a consideration of hydrochemical effects on many aspects of stream ecology. For example, the survival of salmonid eggs in hyporheic gravels varied considerably among the locations studied and was shown to be associated with variation in interstitial chemistry. River restoration schemes and watershed management strategies based only on the surface expression of catchment characteristics risk excluding consideration of potentially critical subsurface processes. Copyright © 2002 John Wiley & Sons, Ltd.     

Field assessment of surface water–groundwater connectivity in a semi‐arid river basin (Murray–Darling,Australia)

In semi‐arid and arid river basins, understanding the connectivity between rivers and alluvial aquifers is one of the key challenges for the management of groundwater resources. The type of connection present (gaining, losing‐connected, transitional and losing‐disconnected) was assessed at 12 sites along six Murray–Darling Basin river reaches. The assessments were made by measuring the hydraulic head in the riparian zone near the rivers to evaluate if the water tables intersected the riverbeds and by measuring fluid pressure (ψ) in the riverbeds. The rationale for the latter was that ψ will always be greater than or equal to zero under connected conditions (either losing or gaining) and always lesser than or equal to zero under losing‐disconnected conditions. A mixture of losing‐disconnected, losing‐connected and gaining conditions was found among the 12 sites. The losing‐disconnected sites all had a riverbed with a lower hydraulic conductivity than the underlying aquifer, usually in the form of a silty clay or clay unit 0.5–2 m in thickness. The riparian water tables were 6 to 25 m below riverbed level at the losing‐disconnected sites but never lower than 1 m below riverbed level at the losing‐connected ones. The contrast in water table depth between connected and disconnected sites was attributed to the conditions at the time of the study, when a severe regional drought had generated a widespread decline in regional water tables. This decline was apparently compensated near losing‐connected rivers by increased infiltration rates, while the decline could not be compensated at the losing‐disconnected rivers because the infiltration rates were already maximal there. Copyright © 2012 John Wiley & Sons, Ltd.     

Groundwater–surface‐water interactions in a braided river: a tracer‐based assessment

Natural tracers (alkalinity and silica) were used to infer groundwater–surface‐water exchanges in the main braided reach of the River Feshie, Cairngorms, Scotland. Stream‐water samples were collected upstream and downstream of the braided section at fortnightly intervals throughout the 2001–2002 hydrological year and subsequently at finer resolution over two rainfall events. The braided reach was found to exert a significant downstream buffering effect on the alkalinity of these waters, particularly at moderate flows (4–8 m³ s^?1/?Q_30–70). Extensive hydrochemical surveys were undertaken to characterize the different source waters feeding the braids. Shallow groundwater flow systems at the edge of the braided floodplain, recharged by effluent streams and hillslope drainage, appeared to be of particular significance. Deeper groundwater was identified closer to the main channel, upwelling through the hyporheic zone. Both sources contributed to the significant groundwater–surface‐water interactions that promote the buffering effect observed through the braided reach. Their impact was less significant at higher flows (>15 m³ s^?1/>Q₁₀) when acidic storm runoff from the peat‐covered catchment headwaters dominated, as well as under baseflow conditions (<4 m³ s^?1/<Q₇₀), when upstream alkalinity was already buffered owing to headwater groundwater sources assuming dominance. The significant temporally and spatially dynamic influence of these groundwater–surface‐water interactions was therefore seen to have important implications for both catchment functioning and instream ecology. Copyright © 2004 John Wiley & Sons, Ltd.     

Coupled dynamic elastic–plastic analysis of earth structures

A fully coupled finite element code based on mixture theory is developed. Prévost's multi-surface constitutive model is tailored to three-dimensional loads and used to predict effective stresses. A new viscous boundary is implemented to avoid wave reflections towards the structure. In contrast to traditional methods, this boundary is able to absorb the two dilatational waves and the shear wave.Two soil deposits and two dams, with different slopes, composed by loose and dense sands have been subjected to the Pacoima accelerogram. Results show how the liquefaction propagates in the soil deposits and earth dams. The importance of the coupling between dilatancy–contractancy and filtration is highlighted by a parametric investigation. Phenomena such as liquefaction and cyclic mobility are reproduced, indicating the robustness of the constitutive model and finite element simulations. As an outcome of the parametric analysis, the seismic stability of dams cannot be improved by decreasing the upstream or downstream slopes.     

Use of non-linear filtering for the regional–residual separation of potential field data

Regional–residual separation is essential in gravity and magnetic data interpretation and a variety of techniques have been proposed. Graphical determination of the regional allows geological information to be taken into account. Upward continuation can be used to obtain the regional field either empirically or using some hypothesis about the geology. In some cases, a matched filter can be designed and used to separate deep and shallow sources. Simple low pass filtering has also been used but without much success. Here we propose to use a non-linear filter approach to remove gravity and magnetic anomalies smaller than a given width. This technique attempts to mimic the graphical separation method. The results from synthetic models are presented as well as the results from a case study in eastern Canada and compared to regional gravity and magnetic anomalies obtained by other techniques. Contrary to the regional fields obtained by upward continuation, non-linear filtering does not have any physical meaning. However, its main advantage is that it gives a regional component of the gravity or magnetic field similar to the one obtained from a graphical separation.     

A hydrobiological study to interpret the presence of desmids in Lake Zirahuén, México

Monthly collections of phytoplankton were supported by physical–chemical data and measures of chlorophyll a concentrations in the search for particular environmental factors that could explain the constant presence of desmids in a Mexican tropical lake, a characteristic not common among the phytoplanktic communities of Mexican lakes. Samplings were taken from the water column in the deepest part of the lake (40 m) and intensified in the metalimnetic zone, whose establishment was monitored by observations of temperature and oxygen profiles. The general behavior of Lake Zirahuén was typical of warm monomictic tropical lakes at high elevation: a short mixing phase during the hemispheric winter. The depth of Z_MIX and Z_eu revealed a well-illuminated epilimnion, suggesting that phytoplankton communities are not likely to be light-limited. The oligotrophic nature of the lake is indicated by discrete concentrations of inorganic nutrients, P_TOT in the interval of 0.01–0.03 mg l⁻¹ and chlorophyll a between 0.23 and 3.98 μg l⁻¹. These characteristics together with a low concentration of calcium, define a lacustrine environment different from other Mexican lakes, and one that could be suitable for desmids communities.     

Why permafrost rocks become unstable: a rock–ice‐mechanical model in time and space

In this paper, we develop a mechanical model that relates the destabilization of thawing permafrost rock slopes to temperature‐related effects on both, rock‐ and ice‐mechanics; and laboratory testing of key assumptions is performed. Degrading permafrost is considered to be an important factor for rock–slope failures in alpine and arctic environments, but the mechanics are poorly understood. The destabilization is commonly attributed to changes in ice‐mechanical properties while bedrock friction and fracture propagation have not been considered yet. However, fracture toughness, compressive and tensile strength decrease by up to 50% and more when intact water‐saturated rock thaws. Based on literature and experiments, we develop a modified Mohr–Coulomb failure criterion for ice‐filled rock fractures that incorporates fracturing of rock bridges, friction of rough fracture surfaces, ductile creep of ice and detachment mechanisms along rock–ice interfaces. Novel laboratory setups were developed to assess the temperature dependency of the friction of ice‐free rock–rock interfaces and the shear detachment of rock–ice interfaces. In degrading permafrost, rock‐mechanical properties may control early stages of destabilization and become more important for higher normal stress, i.e. higher magnitudes of rock–slope failure. Ice‐mechanical properties outbalance the importance of rock‐mechanical components after the deformation accelerates and are more relevant for smaller magnitudes. The model explains why all magnitudes of rock–slope failures can be prepared and triggered by permafrost degradation and is capable of conditioning long para‐glacial response times. Here, we present a synoptic rock‐ and ice‐mechanical model that explains the mechanical destabilization processes operating in warming permafrost rocks. Copyright © 2013 John Wiley & Sons, Ltd.     

Effects of atmospheric oscillations on the field-aligned ion motions in the polar F-region

The field-aligned neutral oscillations in the F-region (altitudes between 165 and 275 km) were compared using data obtained simultaneously with two independent instruments: the European Incoherent Scatter (EISCAT) UHF radar and a scanning Fabry-Perot interferometer (FPI). During the night of February 8, 1997, simultaneous observations with these instruments were conducted at Tromsø, Norway. Theoretically, the field-aligned neutral wind velocity can be obtained from the field-aligned ion velocity and by diffusion and ambipolar diffusion velocities. We thus derived field-aligned neutral wind velocities from the plasma velocities in EISCAT radar data. They were compared with those observed with the FPI (=630.0 nm), which are assumed to be weighted height averages of the actual neutral wind. The weighting function is the normalized height dependent emission rate. We used two model weighting functions to derive the neutral wind from EISCAT data. One was that the neutral wind velocity observed with the FPI is velocity integrated over the entire emission layer and multiplied by the theoretical normalized emission rate. The other was that the neutral wind velocity observed with the FPI corresponds to the velocity only around an altitude where the emission rate has a peak. Differences between the two methods were identified, but not completely clarified. However, the neutral wind velocities from both instruments had peak-to-peak correspondences at oscillation periods of about 10–40 min, shorter than that for the momentum transfer from ions to neutrals, but longer than from neutrals to ions. The synchronizing motions in the neutral wind velocities suggest that the momentum transfer from neutrals to ions was thought to be dominant for the observed field-aligned oscillations rather than the transfer from ions to neutrals. It is concluded that during the observation, the plasma oscillations observed with the EISCAT radar at different altitudes in the F-region are thought to be due to the motion of neutrals.     

Effusive history of the Grande Découverte Volcanic Complex, southern Basse-Terre (Guadeloupe, French West Indies) from new K–Ar Cassignol–Gillot ages

The Grande Découverte Volcanic Complex (GDVC), active since at least 0.2 Ma, is the most recent volcanic complex of the Basse-Terre Island (Guadeloupe, Lesser Antilles Arc). A detailed geochronological study using the K–Ar Cassignol–Gillot technique has been undertaken in order to reconstruct the history of effusive activity of this long-lived volcanic system. Twenty new ages permit to suggest that the GDVC experienced at least six main effusive stages, from 200 ka to present time. To the north of the GDVC, the GDS (Grande Découverte–Soufrière volcano) has been active since at least 200 ka, and to the south, the TRMF (Trois-Rivières–Madeleine Field), started to be emplaced 100 ka. Morphological investigations suggest that the whole TRMF volcanism was emitted from vents distinct from the GDS, most probably a large E–W fissure network linked to the Marie-Galante rift. The mean age of 62 ± 5 ka, obtained for the E–W Madeleine–Le Palmiste alignment suggests that a fissure-opening event occurred at that time. However, whole-rock major and trace element signatures are similar for both systems, suggesting that a common complex magma-plumbing system has fed the overall GDVC. We report very young ages for lava flows from the TRMF, which implies that < 10 ka volcanic activity is now identified for both massifs. Although hazards associated with such effusive volcanism are much lower than those associated with potential flank-collapse of the Soufrière lava dome or a magmatic dome eruption with explosive phases within the GDS, the emplacement of relatively large Holocene age lava flows (3–1 × 10⁸ m³) suggests that a revised integrated volcanic hazard assessment for Southern Basse-Terre should now consider the potential for renewed future activity from two Holocene volcanic centers including the TRMF.     

The 1.5–5-day eastward waves in the upper stratosphere–mesosphere as observed by the Esrange meteor radar and the SABER instrument

Data of neutral meridional wind obtained by the meteor radar at Esrange and data of temperature and pressure measured by the Sounding of the Atmosphere using Broadband Emission Radiometry (SABER) instrument on board the Thermosphere–Ionosphere–Mesosphere Energetics and Dynamics (TIMED) spacecraft were studied with respect to a day-to-day atmospheric variability with periods ranging from 1.5 to 5 days. The detailed analysis was carried out for February 2004. Perturbations of the atmospheric parameters at the examined periods appeared mainly as eastward-propagating waves of zonal wavenumbers 1 and 2. We suggested that these waves excited by the jet instability on both flanks of the polar-night jet in the upper stratosphere and mesosphere interact nonlinearly with each other, and this interaction generates secondary waves. The radar observed both primary and secondary waves at mesospheric heights. The data analysis supports this suggestion. Under conditions of weaker instability observed in February 2003 the perturbations of atmospheric parameters of periods ranging from 1.5 to 5 days had smaller amplitudes at heights of the mesosphere than those in February 2004. It was found that the Eliassen-Palm fluxes calculated for the waves generated by the jet instability were mainly downward directed. This result suggests a possible dynamical influence of the mesospheric layers on the lower atmospheric levels.     

Evaluating the functionality and streamflow impacts of explicitly modelling forest–snow interactions and canopy gaps in a distributed hydrologic model

Many plot‐scale studies have shown that snow‐cover dynamics in forest gaps are distinctly different from those in open and continuously forested areas, and forest gaps have the potential to alter the magnitude and timing of snowmelt. However, the watershed‐level impacts of canopy gap treatment on streamflows are largely unknown. Here, we present the first research that explicitly assesses the impact of canopy gaps on seasonal streamflows and particularly late‐season low flows at the watershed scale. To explicitly model forest–snow interactions in canopy gaps, we made major enhancements to a widely used distributed hydrologic model, distributed hydrology soil vegetation model, with a canopy gap component that represents physical processes of snowpack evolution in the forest gap separately from the surrounding forest on the subgrid scale (within a grid typically 10–150 m). The model predicted snow water equivalent using the enhanced distributed hydrology soil vegetation model showed good agreement (R² > 0.9) with subhourly snow water equivalent measurements collected from open, forested, and canopy gap sites in Idaho, USA. Compared with the original model that does not account for interactions between gaps and surrounding forest, the enhanced model predicted notably later melt in small‐ to medium‐size canopy gaps (the ratio of gap radius (r) to canopy height (h) ≤ 1.2), and snow melt rates exhibited great sensitivity to changing gap size in medium‐size gaps (0.5 ≤ r/h ≤ 1.2). We demonstrated the watershed‐scale implications of canopy gaps on streamflow in the snow‐dominated Chiwawa watershed, WA, USA. With 24% of the watershed drainage area (about 446 km²) converted to gaps of 60 m diameter, the mean annual 7‐day low flow was increased by 19.4% (i.e., 0.37 m³/s), and the mean monthly 7‐day low flows were increased by 13.5% (i.e., 0.26 m³/s) to 40% (i.e., 1.76 m³/s) from late summer through fall. Lastly, in practical implementation of canopy gaps with the same total gap areas, a greater number of distributed small gaps can have greater potential for longer snow retention than a smaller number of large gaps.     

Hydrochemical variations during flood pulses in the south‐west China peak cluster karst: impacts of CaCO3–H2O–CO2 interactions

High‐resolution measurements of rainfall, water level, pH, conductivity, temperature and carbonate chemistry parameters of groundwater at two adjacent locations within the peak cluster karst of the Guilin Karst Experimental Site in Guangxi Province, China, were made with different types of multiparameter sonde. The data were stored using data loggers recording with 2 min or 15 min resolution. Waters from a large, perennial spring represent the exit for the aquifer's conduit flow, and a nearby well measures water in the conduit‐adjacent, fractured media. During flood pulses, the pH of the conduit flow water rises as the conductivity falls. In contrast, and at the same time, the pH of groundwater in the fractures drops, as conductivity rises. As Ca²⁺ and HCO₃^? were the dominant (>90%) ions, we developed linear relationships (both r² > 0·91) between conductivity and those ions, respectively, and in turn calculated variations in the calcite saturation index (SI_C) and CO₂ partial pressure (P) of water during flood pulses. Results indicate that the P of fracture water during flood periods is higher than that at lower flows, and its SI_C is lower. Simultaneously, P of conduit water during the flood period is lower than that at lower flows, and its SI_C also is lower. From these results we conclude that at least two key processes are controlling hydrochemical variations during flood periods: (i) dilution by precipitation and (ii) water–rock–gas interactions. To explain hydrochemical variations in the fracture water, the water–rock–gas interactions may be more important. For example, during flood periods, soil gas with high CO₂ concentrations dissolves in water and enters the fracture system, the water, which in turn has become more highly undersaturated, dissolves more limestone, and the conductivity increases. Dilution of rainfall is more important in controlling hydrochemical variations of conduit water, because rainfall with higher pH (in this area apparently owing to interaction with limestone dust in the lower atmosphere) and low conductivity travels through the conduit system rapidly. These results illustrate that to understand the hydrochemical variations in karst systems, considering only water–rock interactions is not sufficient, and the variable effects of CO₂ on the system should be evaluated. Consideration of water–rock–gas interactions is thus a must in understanding variations in karst hydrochemistry. Copyright © 2004 John Wiley & Sons, Ltd.