Lake Level and Groundwater-Surface-Water Connectivity Assessment

Name the decision owner, geographic and depth domain, time window, evidence use, acceptable uncertainty, and the stage at which the current project must stop. A baseline lake-level description, groundwater-connectivity screening, exchange estimate, private-well concern, septic-source investigation, regulatory determination, mitigation project, and effectiveness study are not interchangeable scopes.

Preserve the ladder below in the project register. Each transition needs a dated record of inputs, assumptions, uncertainty, reviewer, authority, and unresolved alternatives. Field staff should not silently promote an observation into cause, a model estimate into a regulatory finding, or a possible source into public blame.

Sources: [3], [1], [14]

Write the waterbody boundary, connected wetlands and channels, groundwater domain, reference period, and reporting time step. A conceptual balance may include precipitation on the lake, evaporation, measured and unmeasured surface inflow, surface outflow, managed releases, withdrawals and returns, groundwater inflow and outflow, and change in lake storage. Add snow, ice, bank storage, regulation, or other terms when they are material to the setting.

The balance is a bookkeeping framework, not a universal formula that makes every term observable. Align units and time support, retain sign conventions, and label measured, modeled, transferred, assumed, and residual terms. A residual combines omitted processes, bias, timing mismatch, geometry error, and measurement error; it is not automatically groundwater exchange.

Create competing hypotheses before collecting confirmatory data. Examples include climate-driven stage change, altered inflow or outlet control, withdrawals or operations, bathymetric or datum discontinuity, bidirectional groundwater exchange, bank storage, shoreline drainage, septic or sewer leakage, agricultural or wildlife inputs, and more than one mechanism acting together. State the observations that could support, weaken, or leave each hypothesis unresolved.

• Decision, owner, domain, interval, comparison, and stopping point are named.
• Control volume and every candidate inflow, outflow, withdrawal, return, and storage term are diagrammed.
• Measured, modeled, transferred, assumed, and residual values are labeled with units and uncertainty.
• Competing hypotheses and disconfirming evidence are recorded before interpretation.
• A residual is reported as residual or closure error, not relabeled groundwater without independent evidence.

Sources: [2], [3], [1]

Assign a stable site identifier and distinguish gage height from water-surface elevation. Record the gage zero, vertical datum and realization when known, units, benchmark or reference-mark identifiers, survey method, sensor type and position, staff-gage relation, time zone, logging interval, and all moves, resets, replacements, datum corrections, and rating or processing versions.

Tie the lake gage and groundwater measuring points to a common controlled vertical reference through qualified survey work suitable for the intended decision. Protect independent reference marks and preserve level notes. An arbitrary local datum can support one continuous record when its history is controlled, but do not merge it with another gage or well network merely because the numbers look similar.

At each visit, document as-found and as-left readings, independent reference observations, wave or seiche conditions, ice, sensor fouling or movement, vent and barometric status where relevant, clock offset, battery and logger status, and the reason for any correction. Archive raw values before applying approved corrections. Do not convert an ordinary high-water level, operating band, historical statistic, legal elevation, habitat preference, or one community's benchmark into a universal target.

1. 01

   Establish identity and reference

   Record site, gage zero, vertical datum, reference marks, survey lineage, sensor position, units, time basis, and ownership.

2. 02

   Verify in the field

   Compare the logger or sensor to an independent reference and capture conditions, as-found status, maintenance, and as-left status.

3. 03

   Qualify the time series

   Flag gaps, drift, shifts, ice, wave effects, moves, datum changes, provisional corrections, and intervals that cannot be reconciled.

4. 04

   Preserve raw and corrected records

   Keep immutable raw data and a versioned correction ledger with reviewer, rationale, inputs, and changed outputs.

Sources: [5], [13], [1]

Record the bathymetric survey date, sounding and position methods, vertical and horizontal reference systems, stage at survey, shoreline and upland elevation sources, interpolation, grid or triangulation settings, void handling, nearshore coverage, quality checks, and uncertainty. A legacy contour map may be adequate for orientation yet too coarse for a small storage-change estimate.

Build or adopt a versioned stage-area-volume relation that covers the observed stage range and reports extrapolation. Shoreline wetlands, disconnected basins, low-gradient shelves, channels, outlet works, sediment accumulation, dredging, erosion, deposition, vegetation, and changed structures can alter area or storage. The same stage does not guarantee the same volume after the basin or control changes.

Do not interpret an image-derived shoreline as bathymetry or a single depth survey as a timeless storage curve. When the expected storage change is similar to or smaller than stage, survey, interpolation, or shoreline uncertainty, report the interval as unresolved at that resolution rather than forcing a gain or loss.

Sources: [1], [2], [5]

For precipitation, retain station or grid identity, location, elevation, collection interval, time convention, quality flags, gauge exposure, missing-data treatment, snow or mixed-precipitation handling, and the transfer method from observation to lake surface. NOAA station and climate products are useful inputs, but a distant or coarse product is not automatically precipitation on the lake.

For evaporation, state whether the value is measured, modeled, pan-adjusted, energy-budget, mass-transfer, or transferred from another product; retain meteorological inputs, open-water and ice assumptions, spatial footprint, time aggregation, and uncertainty. For surface inflow and outflow, preserve station identity, stage and discharge methods, rating versions, backwater or ice qualifications, ungaged-area treatment, diversions, return flows, outlet operations, and time integration.

Obtain withdrawal, pumping, managed release, augmentation, transfer, and return-flow records from the responsible owner and preserve metering, estimates, operational states, start and stop times, units, confidentiality, and gaps. Align all terms to the same control volume and time grid before subtraction. Do not prescribe a universal monitoring frequency or pumping rate; choose resolution from the decision, response time, variability, detection objective, safety, and approved method.

• Every term has identity, method category, spatial footprint, time support, units, sign, quality status, and uncertainty.
• Station and gridded climate data retain distance, elevation, cell size, quality flags, and transfer assumptions.
• Flow records retain rating or model versions, gaps, ice or backwater qualification, and ungaged-flow treatment.
• Withdrawal, release, augmentation, and return-flow records retain owner, meter or estimate basis, operating times, and confidentiality.
• Budget arithmetic uses aligned intervals and propagates or otherwise transparently carries uncertainty.

Sources: [7], [2], [6]

Select evidence from the conceptual model and expected scale. Water levels in existing wells or authorized piezometers can support head and gradient interpretation when construction, screen interval, aquifer, pumping status, measuring point, survey, and timing are known. Seepage meters estimate exchange across a small bed footprint and interval but can be affected by sediment disturbance, currents, waves, bed gas, installation, leakage, and spatial heterogeneity.

Use temperature, specific conductance, major-ion or other chemistry, environmental isotopes, and natural or introduced tracers only as method categories at planning stage. They may identify contrasts, pathways, mixing, travel-time information, or exchange patterns, but interpretation depends on end members, transformation, season, density, background variability, analytical method, and a qualified specialist. This guide provides no tracer selection, quantity, injection, sampling schedule, or deployment instruction.

Triangulate methods with independent failure modes. Agreement among a budget residual, head pattern, bed-temperature anomaly, seepage measurement, and chemical contrast may strengthen an exchange interpretation; disagreement is diagnostic and must remain visible. A local seepage result is not automatically a whole-lake flux, and no method establishes a septic or contaminant source by itself.

Sources: [4], [3], [1]

Define the target population and sampling frame before choosing accessible locations. Nest lake basins and connected wetlands, shoreline sectors, inflow and outlet zones, developed and reference contexts, transects perpendicular to shore, bed locations, shallow and deeper screen intervals, and lake stations as the conceptual model requires. Record every planned unit, selected unit, replacement rule, access outcome, actual location, and achieved effort.

Nest time around the response process: lake-stage dynamics, rainfall or snowmelt, outlet operations, pumping, irrigation, seasonal groundwater levels, thermal conditions, septic use, and expected travel or mixing lags. Synchronized observations support comparison, while deliberately lagged observations can test a pathway. Select intervals from the decision and process; there is no universal sampling frequency.

Predeclare contrast and replication logic, minimum usable evidence, stopping rules, missing-data treatment, and what would remain non-identifiable. Convenience shoreline sampling can find a feature but cannot support lake-wide prevalence. A dry piezometer, inaccessible parcel, failed logger, frozen site, non-pumping well, or unobserved storm is a method or frame condition, not evidence of no connectivity or no source.

• Target population, frame, selection, nesting, contrasts, replication, and replacement rules are frozen before mobilization.
• Shoreline sector, basin, transect, distance from shore, bed or screen interval, and connected-feature context are retained.
• Timing is related to stage, precipitation, snowmelt, operations, pumping, use, temperature, and plausible lags.
• Planned, completed, partial, inaccessible, unsafe, failed, dry, and out-of-domain units remain distinct.
• The design states which local results may be mapped, averaged, scaled, or left local.

Sources: [1], [4], [6]

For each well or piezometer, retain the measuring point, land-surface and casing reference, vertical datum, screen or open interval, aquifer or unit, construction record, condition, access method, recent pumping or recovery status, water density or temperature considerations when material, measurement method, time, and uncertainty. Convert depth-to-water readings and lake stage to elevations on the same vertical reference before calculating differences.

State the two points, horizontal and vertical separation, times, sign convention, calculation, and propagated or bounded uncertainty for every gradient. Compare uncertainty in the head difference with the observed difference. If datum, survey, clock, lake seiche, barometric response, density, pumping, or screen-interval effects can change the sign, report direction as unresolved or conditional.

A higher groundwater head than contemporaneous lake stage can support a potential gradient toward the lake at that location and interval; the reverse can support a potential gradient away. Heterogeneous sediment, anisotropy, nested flow systems, shoreline geometry, bank storage, delayed response, focused discharge, and seasonal reversal can make a two-point gradient unrepresentative. Direction does not supply hydraulic conductivity, exchange area, or flux, and it does not prove a contaminant reached the lake.

Sources: [6], [4], [5]

Build a controlled source inventory from responsible agency and owner records: permitted and historic wells, screen or completion information where releasable, abandoned wells, septic permits and as-builts, inspection and service history, reported failures, sewers, lift stations, force mains, drains, treatment assets, shoreline drainage, land use, animal or wildlife activity, spills, flooding, soils, and topography. Record record-system owner, access date, completeness, positional accuracy, status, and confidentiality. An absent map feature may mean absent, unrecorded, mislocated, redacted, or outside the queried system.

Use a sanitary survey to record weather, hydrology, waterbody condition, recreational use, visible discharges, odor, turbidity, algal or sewage indicators, wildlife, drainage, infrastructure condition, and candidate sources consistently. Combine it with groundwater direction, plausible travel and transformation, source timing, lake mixing, and alternative pathways. NRCS soil information can support screening of mapped soil and landscape context but does not replace site characterization, septic records, or jurisdictional determinations.

Select analyte, pathogen-indicator, microbial-source, nutrient, major-ion, or other evidence through a project-specific data-quality and laboratory plan. Nitrogen species, phosphorus, chloride, specific conductance, fecal indicators, pathogens, pharmaceuticals, isotopes, or other markers may support different hypotheses, but natural background, agriculture, wildlife, road salt, plumbing, sediment, runoff, dilution, decay, and analytical interference can create overlapping signals. One detection does not identify a septic system; one non-detection does not clear a source.

Treat private drinking-water results as health-sensitive individual information. Use the approved certified laboratory and health-authority pathway, notify the owner and responsible health program as required, and separate household protection from the lake-wide causal investigation. This guide sets no universal septic separation, well construction, test panel, threshold, treatment, or corrective action.

Sources: [9], [10], [11], [8], [3]

Concentration is mass per water volume at a sample and time. Water flux is water volume per area per time, while a flow rate is water volume per time through a defined section or system. Mass flux combines concentration with water flux at a compatible location and interval. Load is mass per time and generally requires representative concentration and flow integration across the relevant pathway and time.

Define whether the calculation concerns lake inflow, outlet export, groundwater exchange, bed-area mass flux, shoreline drainage, septic discharge estimate, withdrawal, or storage change. Match units, spatial support, depth, time, flow direction, censored results, and uncertainty before multiplying. An instantaneous concentration multiplied by an annual flow is not automatically an annual load.

Mixing, dilution, sorption, settling, resuspension, biological uptake, decay, volatilization, transformation, groundwater residence, and lake residence can separate source release from receiving-water concentration. A budget may estimate where mass enters or leaves without uniquely identifying who or what produced it. Report source as a hypothesis until independent source, pathway, timing, and alternative-cause evidence support attribution.

Sources: [2], [4], [3]

For imagery, lidar, elevation, parcel, hydrography, soil, well, septic, utility, shoreline, wetland, bathymetric, land-use, and infrastructure data, record source agency, dataset and layer name, stable identifier, acquisition and access dates, edition or collection, coordinate reference system, vertical datum, units, resolution or scale, accuracy, license or use condition, and the exact downloaded artifact.

Retain scene, band, level, quality mask, cloud, shadow, snow, ice, vegetation, turbidity, water-level, classification, threshold, training, validation, and manual-edit information for remotely sensed shorelines or surface conditions. Landsat Collection 2 supplies controlled products and metadata, but pixel size, water-edge ambiguity, known issues, and acquisition timing still limit parcel-scale or narrow-shoreline conclusions.

Version every transformation: reprojection, datum conversion, clipping, buffering, snapping, interpolation, geocoding, redaction, joining, field calculation, model run, and cartographic generalization. Preserve positional uncertainty and ground checks. A mapped overlap is not proof of connection, ownership, discharge, violation, or cause.

Minimize collection and display of exact private wells, septic components, utility nodes, drinking-water results, cultural resources, Tribal knowledge, access points, and vulnerable infrastructure. Create a restricted analysis layer and a separately reviewed public product using aggregation, masking, coordinate removal, or other approved controls; visual obscuring alone may not prevent re-identification.

• Source, acquisition, edition, identifier, CRS, vertical datum, units, scale or resolution, accuracy, license, and file are retained.
• Remote-sensing scene, processing level, bands, masks, known issues, water condition, classification, validation, and edits are versioned.
• Every reprojection, conversion, join, interpolation, model, and manual correction is reproducible.
• Field locations and mapped features carry positional uncertainty and identity reconciliation status.
• Restricted and public layers have distinct access, redaction, review, retention, and release controls.

Sources: [12], [8], [13]

Use a project quality plan to state objectives, responsibilities, acceptance criteria, calibration and maintenance, reference checks, survey control, field duplicates, blanks and other controls where applicable, sample handling and custody, laboratory methods, detection and reporting conventions, data validation, correction authority, software and model versioning, audits, deviations, and release status. Tailor controls to the decision and method rather than copying a generic checklist.

For water levels, reconcile gage, staff, benchmark, measuring point, screen interval, logger clock, barometric or vent status, and maintenance. For water quality, preserve bottle and lot, location and depth, collection time, preservation, temperature and holding evidence, custody, preparation, method, batch controls, qualifiers, censoring, and laboratory revisions. For records and GIS, preserve query, access date, joins, coordinate transforms, confidence, conflicts, and redaction.

Create a comparability matrix before combining old and new records. Compare site and depth, datum, unit, sensor, method, calibration, interval, time zone, reporting convention, laboratory fraction and method, detection capability, flow integration, bathymetric version, model structure, and quality status. A statistically continuous chart can conceal a method break; mark discontinuities and analyze them separately when harmonization is not defensible.

Freeze immutable raw files and controlled identifiers before editing. Every correction should retain original value, changed value, reason, evidence, author, reviewer, time, affected derivatives, and supersession status. Public release should distinguish raw, corrected, provisional, validated, modeled, estimated, censored, and not-collected values.

1. 01

   Translate the decision into quality objectives

   Define which errors could change the decision and set method-specific acceptance, review, and escalation rules.

2. 02

   Control field and laboratory evidence

   Use approved reference checks, calibrations, survey control, blanks, duplicates, custody, batch review, and corrective action.

3. 03

   Test comparability before merging

   Review datum, site, interval, method, units, laboratory, censoring, geometry, model, and quality status across datasets.

4. 04

   Version corrections and release

   Preserve raw evidence, attributable corrections, reproducible calculations, validation decisions, redaction, and approved public status.

Sources: [14], [15], [5], [6]

Build an authority matrix for lake and launch access; private-property entry; well access and sampling; groundwater-level records; drilling, driving, boring, probing, excavation, and underground-utility clearance; sediment or water collection; seepage devices; tracer work; drones and photography; parks or protected lands; Tribal lands, waters, rights, resources, data, and consultation; health notifications; laboratory release; and publication. Record responsible authority, exact scope and footprint, conditions, dates, personnel, deliverables, stop-work contact, and revocation. Landowner permission does not replace well, utility, environmental, health, Tribal, or agency authority.

Use the employer's approved task- and site-specific safety plans for traffic, boats, weather, lightning, heat and cold, contaminated water or soil, biological hazards, sharps, unstable banks, excavations, utilities, lifting, lone work, communication, decontamination, and emergency response. Never treat a well pit, vault, tank, manhole, or similar enclosure as ordinary access; identify and control confined-space hazards through the responsible occupational-safety program. Ice work requires separate authorization, local assessment, trained personnel, rescue capability, and stop criteria.

Engage the relevant Tribal government or designated office early when work may affect Tribal lands, waters, treaty or similar rights, cultural resources, culturally important species, sacred places, or sensitive knowledge. Government-to-government consultation is performed by the responsible government authority; notification by a consultant, lake group, or data request is not consultation. Follow Tribal direction for access, methods, data ownership, review, storage, and release.

Protect names, addresses, parcel associations, exact private wells and septic systems, household test results, health information, access routes, utility nodes, vulnerable infrastructure, cultural resources, and sensitive Tribal data. Limit collection to the approved purpose; separate identity from analysis; apply role-based access, secure transfer, retention, correction, incident, consent, redaction, and release rules. Public maps and narratives should use the least specific geography that still answers the approved question.

Report non-detection within the sampled matrix, location, depth, time, method, detection capability, recovery, achieved effort, and hydraulic context. Distinguish not detected, not observed, not sampled, inaccessible, unsafe, dry, failed, below reporting level, censored, and outside scope. Close with what the evidence supports, what it weakens, what remains unresolved, plausible alternatives, sensitivity to assumptions, and which later ladder stages require a new decision or authorization.

• Property, water, well, utility, ground-disturbance, collection, device, tracer, drone, park, Tribal, health, laboratory, and publication authorities are resolved separately.
• Confined spaces, underground utilities, ice, boats, weather, unstable ground, contamination, decontamination, and emergency response are covered by approved programs.
• Private drinking-water, exact infrastructure, parcel-linked, Tribal, cultural, and access information has minimization, role, retention, review, and release controls.
• Every planned unit and record is reconciled as complete, partial, failed, inaccessible, unsafe, not collected, censored, or out of scope.
• The final package keeps causal uncertainty, alternatives, method limits, provisional status, corrections, and authority boundaries visible to its audience.

Sources: [15], [18], [17], [16], [11], [14]