Lake-Level, Groundwater, and Source Investigation Program

Name the waterbody and connected system, study boundary, decision owner, data owner, land and infrastructure owners, access authority, water-right and withdrawal context, responsible federal, Tribal, state, territorial, and local programs, designated and existing uses, treaty or reserved rights, cultural and subsistence interests, downstream users, private-well receptors, public-health authority, and communication authority. Record which facts are verified, which are reported, and which remain unresolved. Ownership of shoreline land, a water-control structure, monitoring data, a well, or a septic system does not by itself establish ownership of water, a right to divert it, or authority over another property or jurisdiction.

Translate low lake, suspected groundwater loss, or possible septic influence into a bounded objective: location, period, receptor, proposed hydrologic or source mechanism, evidence needed, uncertainty, alternative causes, adverse endpoints, decision threshold owner, and next authorized decision. EPA's designated-use materials explain the role of uses in water-quality standards, while EPA's Tribal tools recognize that authorized Tribes establish their own uses and may include cultural or traditional purposes. The Meskwaki Nation code is a specific example, not a rule for another Tribe or waterbody. California's water-right process is likewise a jurisdiction-specific example showing why a physical diversion, application, permit, and license are not interchangeable; the governing jurisdiction must be checked directly.

Sources: [2], [3], [4], [18]

Define the lake and contributing groundwater and surface-water boundaries, elevation or storage state, analysis period, time step, seasonal and event windows, connected waters, control structures, withdrawals and returns, and whether each term is measured, estimated, modeled, reported, assumed, or unresolved. Organize precipitation on the lake, surface inflow, surface outflow, groundwater inflow and outflow, evaporation, withdrawals and returns, and change in storage with units, sign convention, coverage, provenance, uncertainty, review status, and dependencies. Use the existing LakeTech water-balance guide for the governing equation and residual cautions.

Distinguish a conceptual model from a computed balance and both from a calibrated groundwater or integrated model. A nonzero closure residual can contain measurement error, datum or area error, temporal mismatch, ungaged flow, storage estimation error, unrecorded operations, unresolved withdrawals, and other omitted processes. It may motivate new evidence, but it does not quantify groundwater exchange or identify a responsible source. Version the conceptual diagram and hypothesis register whenever boundaries or terms change.

• Decision, boundary, period, time step, datum, units, and sign convention are explicit
• Every budget term has a method, owner, source record, uncertainty, coverage, and review state
• Measured, estimated, modeled, reported, assumed, and unresolved values remain visibly distinct
• Storage-area relation, connected waters, operations, withdrawals, returns, and missing terms are documented
• Closure residual is reported with uncertainty and is not relabeled as groundwater exchange
• Conceptual model, computation, model run, review, and release versions are linked but not conflated

Sources: [5], [6], [1]

For every manual or continuous lake-stage record, preserve station identifier and coordinates, waterbody identity, gage type, measuring point, vertical datum and realization, units, time zone, observation time, reference marks, survey lineage, sensor range and resolution, installation and removal, independent checks, fouling or ice condition, barometric or other corrections, telemetry and clock history, gaps, maintenance, rating or storage transformation, reviewer, and provisional or approved state. Keep raw observations, corrected values, derived elevations, storage estimates, and published trend products as linked but distinct records.

Define trend questions before selecting a baseline, aggregation, comparison years, seasonal window, event exclusion, or statistical method. Compare like hydrologic periods and carry datum shifts, gage relocation, changing lake area, control operations, missing data, and method changes into the interpretation. A shoreline photograph, staff-gage observation, sensor series, modeled elevation, and resident recollection can corroborate context but are not interchangeable measures of lake level.

1. 01

   Establish the reference

   Verify station identity, measuring point, vertical datum, reference marks, survey history, units, and transformation lineage before joining records.

2. 02

   Preserve the measurement chain

   Link instrument configuration, field checks, corrections, maintenance, gaps, reviewers, and provisional or approved status to every interval.

3. 03

   State the trend contrast

   Name the period, baseline, seasonal or event frame, coverage rule, uncertainty, and factors that could create an apparent trend.

4. 04

   Bound the claim

   Report the supported level or storage pattern without assigning a groundwater pathway, withdrawal, source, or remedy from stage alone.

Sources: [9], [5], [1]

Inventory tributaries, channels, culverts, springs expressed at the surface, wetlands, overflows, dams and control structures, diversions, withdrawals, discharges, return flows, stormwater and wastewater infrastructure, and ungaged drainage. For each flow record preserve site geometry, method, stage or discharge basis, rating or computation version, field conditions, control changes, reversals, ice or backwater, coverage, units, reviewer, and whether the result is measured, estimated, reported, or modeled. Do not apply a point measurement to an unrepresented event or period without a documented method and uncertainty.

Match precipitation and evaporation evidence to the lake surface, period, units, gauge exposure or gridded-product support, missing-data rule, phase and snow treatment, and lake-area assumption. Reconcile reported pumping, operational logs, permitted or registered quantities, observed infrastructure, and return pathways without assuming that a permit limit equals actual use or that unlisted equipment proves an unauthorized withdrawal. Where a term remains weak, carry it as uncertainty or design a bounded method to improve it.

Sources: [10], [5], [18]

Create a hydrostratigraphic and infrastructure model before comparing lake stage with well or piezometer levels. Preserve well identity, ownership and access, purpose, coordinates and location confidence, construction source, screened or open interval, aquifer assignment, measuring point, land-surface and vertical datum, static or pumping condition, recent pumping, well integrity, method, instrument checks, barometric treatment where applicable, time synchronization, and uncertainty. Convert levels to a common vertical reference only when datum lineage supports it. A domestic-well level affected by pumping, an unknown open interval, or an imprecise land-surface elevation may not represent the head needed for a lakebed exchange question.

Evaluate exchange at the spatial and temporal scale of the decision using converging evidence categories such as paired heads and gradients, seepage or flux measurements under an approved method, temperature or other physical patterns, geologic and geophysical information, water chemistry and environmental tracers, event responses, and a water-budget consistency check. Each method has scale, heterogeneity, detection, mixing, and interpretation limits. A downward gradient does not prove a contaminant reached a well; a chemical similarity does not prove exchange direction; an environmental tracer can support source or residence interpretation but does not by itself identify a household or authorize an introduced tracer study.

• Hydrostratigraphic units, lakebed setting, well intervals, and possible preferential pathways are explicit
• Lake and groundwater elevations share verified datum, time basis, units, and uncertainty
• Pumping, recovery, barometric, seasonal, event, construction, and integrity effects are documented
• Exchange direction, timing, magnitude, and connectivity are treated as separate questions
• At least one independent evidence category tests the head-based interpretation
• Contradictory and unresolved evidence remains visible in the conclusion

Sources: [6], [7], [8]

Use a repeatable sanitary-survey frame to document shoreline and watershed conditions, tributaries and outfalls, stormwater and wastewater infrastructure, septic indicators, sanitary facilities, animals and wildlife, recreation and occupancy, weather, drainage, erosion, odors, visible discharges, maintenance observations, and public reports. Preserve location, time, observer, method, photograph or record link, access basis, uncertainty, and whether the item is a direct observation, owner statement, agency record, map inference, or unresolved lead. An odor, lush vegetation, algae, pipe, wet area, bacteria result, or resident report is not source attribution.

Request property, parcel, permit, inspection, complaint, repair, septic, well-construction, pumping, water-use, sewer, utility, and historical imagery records from the responsible custodian under applicable access rules. Record source agency, retrieval date, coverage, location confidence, status, gaps, and use restrictions. The Minnesota Well Index is a state-specific example: it provides useful construction and static-level information while warning that records can be absent and unverified locations can be inaccurate. Absence from a database does not prove absence on a property, and one state's records do not apply elsewhere.

Create a restricted crosswalk for parcel, household, well, septic, access contact, exact infrastructure coordinates, and sample identifiers. Publish generalized locations or coded identifiers only when they preserve the authorized decision use. Do not place owner names, contact details, household health information, exact private-well or septic locations, access codes, security-sensitive infrastructure, or unreviewed allegations in the public investigation record.

Sources: [11], [12], [17], [1]

Build a hypothesis-to-evidence matrix before selecting samples. Controlled categories may include field physical and chemical measurements; major ions and general chemistry; nutrients and species; fecal indicator organisms; microbial source-tracking markers; wastewater-associated or other chemical indicators; stable, radioactive, or naturally occurring environmental tracers; temperature and specific conductance patterns; and site-specific constituents identified by the responsible health or regulatory authority. For every category name the question, matrix, station and depth logic, hydrologic condition, analytical method and laboratory, containers and preservation, custody, field and laboratory QA/QC, reporting and detection information, confounders, data-use rule, privacy class, and confirmatory evidence needed.

Microbial source tracking can add host-association evidence, but marker performance, persistence, transport, laboratory contamination, non-detects, cross-reactivity, and local validation constrain the claim. Nutrients, chloride, optical properties, pharmaceuticals, caffeine, artificial sweeteners, isotopes, radon, major ions, or other chemicals can have multiple sources and different transport behavior. Use mixtures and spatial or temporal patterns to test hypotheses; do not treat one detection or correlation as a property-level source identification.

Environmental tracers are naturally occurring constituents, isotopes, or physical properties used to investigate movement and mixing. An introduced dye, salt, gas, microbe, or other tracer is a different activity that may require toxicological review, owner consent, permits, utility or public-health coordination, a recovery and communication plan, and qualified method design. This guide does not specify a tracer, amount, injection, sampling frequency, or release method.

Sources: [15], [6], [1], [14]

Maintain a hypothesis register that includes septic systems, sewer or sanitary infrastructure, stormwater, tributaries, groundwater discharge, private or public wells and pumping, lake operations, agriculture, wildlife and pets, recreation, atmospheric inputs, shoreline erosion, internal cycling, geologic sources, laboratory or field contamination, and data-processing artifacts where relevant. For each hypothesis identify a plausible source, release mechanism, transport path, receptor, timing, expected evidence, evidence against, alternative explanations, and discriminating next observation. Add hypotheses as evidence warrants; do not preserve a preferred source by redefining every contradiction as noise.

Build the attribution case from independent categories: verified infrastructure and records, source-proximal observations, hydraulic or transport feasibility, temporal sequence, spatial pattern, source-associated microbial or chemical evidence, comparative or background locations, repeat observations, and response to an independently documented change. Evaluate data quality, strength, consistency, specificity, temporality, plausibility, coherence, alternative causes, and uncertainty. Co-located septic and lake effects do not prove transport; a head gradient does not identify the solute source; a marker does not prove a specific property; and improvement after a repair can support but may not uniquely prove cause when rainfall, loading, operations, or sampling also changed.

1. 01

   State competing hypotheses

   Give each plausible source and non-source explanation a transport mechanism, expected observations, contradictions, and decision consequence.

2. 02

   Map each evidence item

   Classify provenance, method quality, independence, spatial and temporal support, privacy, and whether the evidence supports, weakens, or does not discriminate a hypothesis.

3. 03

   Seek discriminating evidence

   Prefer observations that separate hypotheses over repeated collection of the same ambiguous indicator.

4. 04

   Document the weight of evidence

   Report support, contradictions, alternatives, uncertainty, reviewer, and the bounded attribution state without making a legal or health determination.

Sources: [16], [15], [11]

Create an authority matrix for recreational-water notices, public drinking-water systems, private wells, wastewater and septic systems, stormwater and discharges, groundwater and water rights, wells and drilling, dam and control structures, protected species and cultural resources, laboratory certification, property access, records, and emergency response. Name the decision, authority, submission, evidence standard, status, conditions, appeal or correction path, and communication owner. Federal private-well materials explain that federal public-system drinking-water rules generally do not regulate private wells and that owners have important responsibilities, but state, Tribal, territorial, county, or local requirements and assistance can differ. The responsible health department or other authority must interpret site-specific health implications and actions.

Obtain documented consent or other lawful authority before entering property, inspecting infrastructure, measuring a private well, collecting household water, interviewing occupants, installing equipment, or sharing property-linked results. Separate consent to access, consent to sample, data-use permission, return-of-results responsibility, emergency contact, and public-release permission. Plan how urgent individual results reach the owner and health authority without waiting for a general report, and how aggregate communication avoids revealing a household through exact coordinates, rare attributes, photographs, small-area maps, or coded tables that can be reidentified.

• Each health, regulatory, access, records, water-right, and communication decision has a named authority
• Private-well ownership responsibility is not confused with public-system regulation or health-department authority
• Property entry, inspection, sampling, installation, interview, and release permissions are separately documented
• Exact infrastructure and household data are restricted by role, purpose, retention, and release rule
• Urgent individual notification and aggregate public communication have distinct workflows
• Investigation language does not declare safety, violation, liability, eligibility, or authorization

Sources: [13], [14], [12], [18]

Compare no action with continued observation, data repair, access or records work, focused hydrologic monitoring, sanitary-survey follow-up, private-well or public-health referral, infrastructure inspection, source control, operation changes, conservation, repair, connection, replacement, watershed measures, or other jurisdiction-defined categories. For each category document the objective, evidence fit, assumptions, uncertainty reduced, target and non-target receptors, potential adverse outcomes, equity and cultural considerations, access and privacy burden, operations and maintenance, reversibility, monitoring need, legal pathway, cost class, and what would trigger reconsideration. No action still needs explicit consequences, monitoring, and reopening criteria.

Record mitigation or funding eligibility as a decision by the named program under its current criteria. Keep it separate from owner consent, water rights, land-use approval, public-health direction, permits, environmental and cultural review, procurement, engineering approval, and authority to work. Likewise, an application, grant award, plan review, or preliminary agency conversation is not final authorization unless the responsible authority's record says so. Do not convert this alternatives screen into a septic setback, well design, pumping plan, remediation, treatment, product, or engineering recommendation.

Sources: [16], [3], [18]

Before work, freeze the approved decision basis, drawings or procedures, locations, access, methods, materials, equipment, personnel qualifications, sequencing, hold points, environmental and cultural protections, public-health controls, monitoring, stop-work criteria, adverse-response plan, reporting, and change authority. Confirm that authorization remains current and applies to the exact property, infrastructure, waterbody, activity, timing, and personnel. Procurement, contractor availability, or equipment capability does not enlarge the approved scope.

During work, preserve dates and times, station and asset identifiers, field logs, photographs under the privacy plan, equipment and material records, calibration or test records, quantities where the approved design requires them, weather and hydrologic context, inspections, samples and custody links, unexpected conditions, stop-work events, complaints, deviations, author and approver, corrective action, and final acceptance. Produce an as-built or as-performed record that shows what changed from the approved plan and whether the change was authorized before it occurred. An invoice, completion statement, or installed device is implementation evidence, not effectiveness evidence.

• The final authorization and approved plan are versioned and linked to the decision basis
• Property, water, infrastructure, sampling, tracer, pumping, construction, and release scope match the authorization
• Hold points, stop-work criteria, adverse-response roles, and change authority are explicit
• Field evidence preserves actual locations, methods, timing, conditions, deviations, and approvals
• Sensitive household, well, septic, and infrastructure records remain in the restricted system
• As-approved, as-performed, as-built, accepted, and operational states are distinct

Sources: [1], [18]

Before implementation, define process endpoints, implementation endpoints, hydrologic endpoints, source-pathway endpoints, water-quality endpoints, private-well and public-health referral endpoints, non-target and downstream endpoints, adverse endpoints, operations and maintenance endpoints, communication endpoints, analysis periods, baseline and comparison rules, uncertainty, decision owner, and response to mixed evidence. Link each endpoint to the mechanism and to data that can actually support the intended inference. Do not insert a universal target level, minimum flow, sample frequency, analyte threshold, setback, separation, or treatment target; the responsible authority and approved site-specific plan supply any governing values.

Evaluate weather, lake operations, watershed loading, withdrawals, groundwater conditions, sampling coverage, laboratory or sensor changes, and concurrent actions before attributing a response. Confirm implementation separately. Then ask whether the expected pathway changed, whether the target receptor responded, whether non-target or adverse outcomes occurred, and whether the result persisted for the bounded period. No detection, no complaint, apparent level recovery, reduced concentration, or a modeled improvement may be useful evidence but does not alone establish absence of exposure, causal success, regulatory compliance, or durability.

Sources: [1], [16], [5]

At each review point, decide whether evidence supports continued monitoring, a method correction, expanded or narrowed hypotheses, authority referral, a different alternative, stop-work, corrective action, or closure. Predefine triggers for datum or instrument failure, access loss, contaminated or compromised samples, unexpected pumping or operations, contradictory exchange evidence, sensitive-data exposure, adverse outcomes, illness or well concerns, permit or authority change, and failed implementation. Preserve the original record, revised value or interpretation, reason, affected scope and period, author, reviewer, approval, downstream products, owner or authority notifications, and reissue status.

Release observation, hydrologic state, exchange evidence, source hypothesis, causal attribution, public-health or regulatory status, mitigation eligibility, authorization, implementation, effectiveness, and communication as separately attributable fields. Publish only the states the release authority approved. State boundary, period, methods, coverage, uncertainty, alternatives, contradictions, privacy treatment, sensitive-location generalization, record date, version, sources, reviewer, and correction channel. A public summary should never reveal exact private-well or septic coordinates, household identities or health information, access details, security-sensitive infrastructure, or unreviewed property-level allegations.

Link the released program record to the planned LakeTech measurement guide, field form, and pathway; the existing water-balance, sampling-plan, sample-to-laboratory, data-validation and release, watershed-load, and management-effectiveness guides; and the restricted source records needed for audit. If evidence changes, issue a dated correction or withdrawal to every affected product and audience rather than silently replacing the result. Close a phase only when unresolved questions, retention and access rules, operation and maintenance ownership, future review triggers, and the next responsible decision are explicit.

1. 01

   Review the evidence state

   Reconcile observations, quality controls, water-budget terms, exchange evidence, hypotheses, authority records, implementation, endpoints, contradictions, and privacy state.

2. 02

   Choose the authorized branch

   Continue, correct, pause, stop, refer, revise the plan, compare another alternative, or close only through the named decision owner and change process.

3. 03

   Preserve corrections

   Retain prior values and releases, reason, affected scope, reviewer, notifications, revised products, and effective date.

4. 04

   Release the bounded record

   Share only approved decision states with limitations, privacy protection, source lineage, version, and correction route.

Sources: [19], [1], [13]