These interactive surfaces map water-column metal concentration (μg/L) across approximately 820 km of the Pilcomayo River Basin and eight years of monitoring campaigns (2016–2024). Space, time, and intensity rendered together in a single view.
As · Cd · Pb · Zn820 km Corridor8 YearsInteractive 3D
Methodology & Interpretive Caveats
How to Read These Surfaces
Each surface plots the same three dimensions: downstream distance from Potosí (Y axis, km), monitoring campaign date (X axis, 2016–2024), and water-column metal concentration in μg/L (Z, vertical height). Three caveats govern interpretation.
⚠
Concurrent snapshots — not parcel trackingThese models represent concurrent spatial measurements across the monitoring network rather than tracking of individual water parcels downstream. A ridge that appears to shift downstream over time is not the same pulse of contamination moving through the basin — it reflects changing concentration levels measured simultaneously at different stations. Interpreting apparent downstream movement as transport is a misreading of this data type.
◈
Distance assignment is non-trivialRiver distances were assigned to each monitoring station based on manually classified hydrological position within the basin, correcting for the complex tributary structure of the upper Pilcomayo. Simple straight-line or single-channel distances from Potosí would misrepresent station positions — particularly in the upper basin where the main channel divides and reconnects across the Bolivian Chaco.
◎
Interpolation & interactionValues between sampling stations and campaigns are linearly interpolated across a 50×50 spatiotemporal grid with nearest-neighbor edge fill — the smooth surface you see is an interpolation, not a continuous measurement. To interact: Rotate — click and drag · Zoom — scroll wheel · Pan — right-click and drag · Hover — point values appear as a tooltip.
Predicted Visual Signatures
What Each Surface Should Reveal
Each surface is a predicted-vs-observed experiment. The partitioning hierarchy established on the Sediment Partitioning page — Pb most sediment-bound, Cd most mobile, As and Zn intermediate — should produce distinct visual signatures across the four surfaces. Here is what to look for in each one.
Pb Lead
1,667 L/kg
Expected surface: narrow, steep source-proximal peak that drops away sharply downstream. The high Kd means Pb is rapidly removed from the water column by sediment adsorption, leaving the surface nearly flat beyond the headwater zone. Elevated values should appear only as sharp wet-season spikes or during hypoxic mobilization events — not as a persistent plume.
Expected surface: strong source-proximal peak with a gradual — rather than abrupt — downstream decay, because Fe/Al-oxide buffering distributes arsenic across multiple reaches before final removal. The report documents more than an order-of-magnitude decline from headwater to distal stations. Wet-season ridges should be visible, though less extreme than Cd.
Expected surface: flatter spatial gradient than Pb or As, because the lower Kd allows Cd to persist in the water column further downstream. The most visually distinctive feature should be in the time dimension: strong wet-season ridges driven by the acid-soluble Ca/P pathway, which releases Cd into solution with wet-season runoff acidity and elevated discharge.
Expected surface: the most spatially homogeneous of the four. Seven verified binding mechanisms working simultaneously — across Fe/Mn oxides, carbonates, organic matter, sulfides, and co-precipitation pathways — give Zn a resilient, distributed sediment affinity that buffers against the sharp peaks and troughs visible in the other metals.
Arsenic & Lead: A Tale of Two Water-Column Signals
Both metals are controlled by Fe and Al oxide phases, but their water-column surfaces are strikingly different. Arsenic shows a persistent, graduated source-to-distal gradient — the report documents water-column concentrations more than an order of magnitude higher at the Potosí headwaters than at distal stations near the Argentina/Paraguay border. Lead, by contrast, is so strongly sequestered by its high Kd (1,667 L/kg) that its water-column surface is largely suppressed under baseline conditions. The water-column signal for Pb concentrates into sharp, transient spikes during wet-season runoff pulses. This is the inverse of the sediment story documented on the Sediment Texture page, where Pb appears persistently elevated near the source — the two pages are telling complementary halves of the same contamination picture.
AsArsenic · Kd 908 L/kg
What you are looking at: Water-column arsenic concentration (μg/L) at every monitoring station across every sampling campaign. The Y axis is downstream distance from Potosí (km); the X axis is time (2016–2024); the Z axis (vertical height) is concentration. The tall ridge along the near edge (km ≈ 0–50) is the Potosí source zone. Moving along the Y axis, the surface should fall sharply — the report documents a decline of more than an order of magnitude from headwater stations to distal stations near the Argentina/Paraguay border. Matrix: water column · Units: μg/L · Interpolation: 50×50 linear grid · Rotate to compare; hover for point values.↗ Open fullscreen
PbLead · Kd 1,667 L/kg
What you are looking at: Water-column lead concentration (μg/L) across space and time. Because Pb's Kd (1,667 L/kg) is the highest of the four tracked metals, most lead is sequestered into sediment almost immediately after entering the water column. The surface is therefore expected to be relatively flat and low under baseline dry-season conditions, with concentrated peaks appearing in the time dimension as wet-season runoff pulses mobilize Pb from disturbed sediment. Acute hypoxic events produce the sharpest spikes; see the Hypoxic Events page for the three documented events. Matrix: water column · Units: μg/L · Interpolation: 50×50 linear grid · Rotate to compare; hover for point values.↗ Open fullscreen
Ca/P and Multi-Site Binding
Cadmium & Zinc: Mobility Written in the Time Dimension
Cadmium (lowest Kd of the four metals at 200 L/kg) and zinc (intermediate at 664 L/kg) both show flatter spatial gradients than lead or arsenic — their weaker sediment binding allows them to travel further downstream in the water column before being removed. The more interesting contrast is in the time dimension: cadmium shows the strongest wet-season pulse signature of any of the four metals, driven by its Ca/P binding pathway, which is acid-sensitive and responds sharply to the elevated discharge and acidity of wet-season runoff. Zinc, by contrast, is the most spatially homogeneous surface — its seven simultaneous binding mechanisms distribute it across the basin without the pronounced peaks that characterize the other metals.
CdCadmium · Kd 200 L/kg
What you are looking at: Water-column cadmium concentration (μg/L) across the basin and monitoring period. With the lowest Kd of the four tracked metals (200 L/kg), cadmium remains in the water column longer than any other metal measured here. Focus especially on the time (X) axis: the wet-season ridges (Nov–Apr) should be the most visually pronounced of all four surfaces, reflecting the acid-soluble Ca/P pathway releasing Cd with each wet-season runoff cycle. Matrix: water column · Units: μg/L · Interpolation: 50×50 linear grid · Rotate to compare; hover for point values.↗ Open fullscreen
ZnZinc · Kd 664 L/kg
What you are looking at: Water-column zinc concentration (μg/L). Of the four surfaces, zinc should present the flattest, most spatially uniform landscape — the visual signature of seven simultaneous binding mechanisms that distribute zinc across sediment phases throughout the basin. There is no single dominant pathway to disrupt, so neither source proximity nor wet-season runoff produces the dramatic peaks that distinguish lead, arsenic, and cadmium. Matrix: water column · Units: μg/L · Interpolation: 50×50 linear grid · Rotate to compare; hover for point values.↗ Open fullscreen
What the Models Show
Patterns Across All Four Metals
Source Zone
Water-Column Saturation at Potosí
All four metals show elevated water-column concentrations within the first 0–50 km of the Potosí headwaters. For arsenic, the report documents wet-season medians of approximately 47.5 μg/L at source-proximal stations, dropping to approximately 15.0 μg/L in the dry season. Lead reaches water-column concentrations of approximately 292 μg/L during wet-season pulses at downstream stations. The complementary sediment story — concentrations at the mg/kg scale at San Antonio–Potosí — belongs on the Sediment Texture page.
Spatial Gradient
Downstream Dilution and Sedimentation
For arsenic — the best-quantified case — the report documents a decline of more than an order of magnitude in water-column concentration from source-proximal stations at Potosí to distal stations at the Argentina/Paraguay border. The gradient is steepest for lead and arsenic; gentler for cadmium and zinc. The binding-mechanism explanation for each gradient's shape is on the Sediment Partitioning page.
Transboundary
Transboundary Persistence
Persistent non-zero concentrations at downstream stations confirm that contamination reaches transboundary waters throughout the monitoring record. This is the data justification for the trinational scope of the Comisión Trinacional and the River Remedy project itself: a monitoring or remediation framework that stops at the national boundary is structurally incomplete.
Temporal Signal
Seasonal Wet-Season Pulse (Nov–Apr)
Wet-season concentration spikes (November–April) are visible as ridges in the time dimension of each surface. These ridges are the visual evidence for the 3× to 30× wet/dry water-column ratios quantified on the Spatiotemporal Migration page. The amplitude varies by metal — cadmium shows the strongest pulse; zinc the most muted contrast.
Methodological Limits
What These Surfaces Can't Show
These surfaces are powerful orientation tools, but they are also abstractions. Five things fall outside what they can represent.
⊟
The sediment reservoir is invisible hereThese surfaces are water-column only. The sediment-bound fraction — which holds the majority of total metal mass in the basin — does not appear. The spatial gradient of sediment contamination is the subject of the Sediment Texture page; the binding framework connecting the two matrices is on the Sediment Partitioning page.
≋
Values between stations are interpolated, not measuredThe smooth surface between monitoring stations is a linear interpolation across a 50×50 grid — a mathematical estimate, not a physical observation. Processes occurring between stations (tributary inputs, point-source discharges, localized sedimentation) are not captured.
◎
Apparent downstream movement is not transportAs noted in 'How to Read These Surfaces' above, any apparent shift in a ridge over time reflects changing concentration levels at fixed stations — not a plume moving through the basin. The surfaces encode concurrent spatial snapshots, not Lagrangian transport.
◷
Acute hypoxic events are smoothed overThe three documented hypoxic mobilization events (detailed on the Hypoxic Events page) involved sharp, short-duration concentration spikes that are likely compressed or averaged at the temporal resolution of these surfaces.
⊕
Mercury and the 2024 emerging contamination event are not mapped here
The 2024 emerging contamination event (Sediment Texture) involves a mercury and cadmium co-spike. Mercury is not among the four modeled metals and does not appear. Whether the 2024 Cd spike is distinguishable in the Cd surface has not been independently verified against the iframe dataset.
Where to Go Next
The Spatial Picture, Complete
These four surfaces are the spatial synthesis of the site — the point where partitioning chemistry, seasonal cycling, and source dominance all become simultaneously visible. Each sister page developed one dimension of this picture in depth.