{"schema_version":1,"experiment_id":"eoa_inverse_innovation_exp11_mechanism_context_external20_20260804","research_id":"eoa_inverse_innovation_exp11_external_scrutiny_20260804","cell_id":"deadweight_loss_reduction__earth_sciences","opaque_id":"deadweight_loss_reduction__earth_sciences__C","search_lanes":{"direct_problem":{"queries":["Earth system model output storage quota blocks runs hot storage tiering archive","climate model data storage quota workflow runs archive tiering HPC","climate modeling simulation storage bottleneck output limits experiments paper","Earth system model storage constraint prevents simulation runs quota"],"source_ids":["SRC1","SRC2","SRC3"],"no_result_note":null},"closest_prior_art":{"queries":["climate simulation output data lifecycle policy scratch campaign tape archive recall quotas","Earth system modeling data management lifecycle hot warm cold storage repository","Community Earth System Model data management policies storage preservation curation","ECMWF weather modelling hierarchical storage disk cache HPSS"],"source_ids":["SRC3","SRC4","SRC5","SRC6"],"no_result_note":null},"historical_terminology":{"queries":["HPC hierarchical storage management automatic migration LRU tape quota scientific data retention policy","hierarchical storage management climate model output archive","meteorological archival retrieval system disk cache tape","policies for data archival hierarchical storage management LRU"],"source_ids":["SRC4","SRC6","SRC7"],"no_result_note":null},"products_practices_standards":{"queries":["official object storage lifecycle tiering access frequency automatic restore retention legal hold checksum provenance","ISO OAIS preservation information fixity provenance access rights official standard","HPC hierarchical storage management automatic migration tape retrieval product","scientific archive storage migration access dissemination standard"],"source_ids":["SRC5","SRC6","SRC7","SRC8"],"no_result_note":null},"non_english_regional":{"queries":["site:dkrz.de Klimamodelldaten Speicher Kontingent Archiv Hierarchie","Klimadaten Speicherhierarchie automatische Migration Band Rückholung Kosten DKRZ","Datenarchiv Klimamodellergebnisse Hochleistungsrechner Bandbibliothek Cache"],"source_ids":["SRC5"],"no_result_note":null},"composition_subproblems":{"queries":["scientific repository storage tiering pilot recall latency cost audit LRU migration","climate simulation output data lifecycle retention provenance recall cost quota","CMIP data retention tiers preservation ESGF replica priority tiers","climate model storage limited usage anticipated demand retention governance"],"source_ids":["SRC2","SRC3","SRC6","SRC7","SRC8"],"no_result_note":null}},"sources":[{"source_id":"SRC1","title":"Evaluating lossy data compression on climate simulation data within a large ensemble","url":"https://www.research-collection.ethz.ch/entities/publication/211d2ce7-d3af-4f0b-a089-deff8c7679d2","publisher":"Copernicus Publications; hosted by ETH Zurich Research Collection","date_or_year":"2016","source_type":"PRIMARY_RESEARCH","language":"English","claims_supported":["High-resolution Earth-system simulations generate volumes that strain institutional storage.","Storage requirements can force reductions in output frequency, simulation length, or ensemble size.","Scientific-integrity testing is necessary when reducing the storage burden."]},{"source_id":"SRC2","title":"Quotas — NERSC Documentation","url":"https://docs.nersc.gov/filesystems/quotas/","publisher":"National Energy Research Scientific Computing Center","date_or_year":"Current documentation, accessed 2026-08-04","source_type":"OFFICIAL_GUIDANCE","language":"English","claims_supported":["HPC filesystems impose fixed space and inode quotas.","Exceeding a quota can prevent new data from being written even though jobs can still be submitted.","NERSC separately reports project filesystem and HPSS archive quotas."]},{"source_id":"SRC3","title":"Data Management and Data Distribution Plan for the CESM Project","url":"https://files.cesm.ucar.edu/management/docs/cesm-data-management-plan-2021.pdf","publisher":"Community Earth System Model Project, National Center for Atmospheric Research","date_or_year":"2022","source_type":"OFFICIAL_GUIDANCE","language":"English","claims_supported":["CESM explicitly states that storage capacity is limited and cannot retain all output indefinitely.","CESM applies different retention strategies by data category, usage, anticipated demand, and scientific value.","The CESM Data Manager, principal investigators, Chief Scientist, and Scientific Steering Committee have defined stewardship and exception authority.","The plan balances preservation and access against storage and labor costs and defines verification before removing raw history output."]},{"source_id":"SRC4","title":"Community Earth System Model Data Management: Policies and Challenges","url":"https://www.sciencedirect.com/science/article/pii/S1877050911001165","publisher":"Elsevier, Procedia Computer Science","date_or_year":"2011","source_type":"PRIMARY_RESEARCH","language":"English","claims_supported":["CESM has long-standing formal data-management policies and strategies.","Climate-model output creates established challenges involving preservation, curation, provenance, metadata, and distribution.","The proposal sits in a mature Earth-system data-management problem area rather than an unrecognized one."]},{"source_id":"SRC5","title":"Datenarchiv","url":"https://www.dkrz.de/de/systeme/datenarchiv","publisher":"Deutsches Klimarechenzentrum GmbH","date_or_year":"2026","source_type":"FIRST_PARTY_PRODUCT","language":"German","claims_supported":["DKRZ archives roughly 30 PB of new climate-model results per year.","Its operational archive combines an 8.2 PB SSD/disk cache with automated tape libraries exceeding 415 PB capacity.","A regional climate-computing center already operates a disk-cache-to-archive architecture for continuous simulation-data flow and retrieval."]},{"source_id":"SRC6","title":"Data Handling System","url":"https://www.ecmwf.int/en/computing/our-facilities/data-handling-system","publisher":"European Centre for Medium-Range Weather Forecasts","date_or_year":"Current page, accessed 2026-08-04","source_type":"FIRST_PARTY_PRODUCT","language":"English","claims_supported":["ECMWF routinely stores and retrieves operational and research weather-model data through a large-scale data-handling system.","MARS stages recent or accessed data in disk caches while storing the bulk in HPSS.","ECFS uses an HPSS disk-and-tape hierarchy while preserving a logical filesystem view and metadata.","The system has identifiable institutional operators and thousands of users."]},{"source_id":"SRC7","title":"Hierarchical Storage Management (HSM)","url":"https://www.ibm.com/docs/en/i/7.5.0?topic=feature-hierarchical-storage-management-hsm","publisher":"IBM","date_or_year":"Current product documentation, accessed 2026-08-04","source_type":"FIRST_PARTY_PRODUCT","language":"English","claims_supported":["HSM automatically distributes data across storage layers to meet access needs while reducing cost.","Migration policies can move selected data from faster expensive storage to slower cheaper storage and promote it when access increases.","Archive eligibility can use inactivity, size, or usage criteria, and access can trigger dynamic retrieval."]},{"source_id":"SRC8","title":"ISO 14721:2025 — Space Data System Practices — Reference model for an open archival information system (OAIS)","url":"https://www.iso.org/cms/%20render/live/en/sites/isoorg/contents/data/standard/08/74/87471.html","publisher":"International Organization for Standardization","date_or_year":"2025","source_type":"OFFICIAL_STANDARD","language":"English","claims_supported":["A responsible archive must preserve information and make it available to a designated community through policy-based processes.","OAIS covers archival storage, data management, access, dissemination, and migration to new media or forms.","Preservation and accessibility are substantive constraints on any storage-tiering intervention."]}],"problem_evidence":{"status":"PARTLY_SUPPORTED","finding":"The general wedge is real: Earth-system simulations strain institutional storage, storage limits can reduce scientific output, and an exceeded HPC quota can prevent new writes. However, the bounded search found no public event log showing a particular repository with four Earth-model runs blocked solely by a flat hot-storage quota while at least 200 TiB of safely tierable inactive output occupied that tier.","source_ids":["SRC1","SRC2","SRC3"],"uncertainty":"Public sources establish storage pressure and quota mechanics, but not the proposal's repository-specific causal count, required bytes, sole-blocker attribution, or lifecycle-cost thresholds."},"adopter_evidence":{"status":"SUPPORTED","finding":"Repository and modeling-program operators are identifiable adopters and authorizers. CESM assigns responsibilities to its Data Manager, investigators, Chief Scientist, and Scientific Steering Committee; DKRZ and ECMWF operate domain-specific archives and hierarchical storage systems serving modelers.","source_ids":["SRC3","SRC5","SRC6"],"uncertainty":"The sources establish analogous institutional authority structures, not approval from the unspecified pilot repository."},"implementation_evidence":{"status":"PARTLY_SUPPORTED","finding":"The core mechanism is implemented routinely: climate and weather centers use disk caches, tape archives, HPSS, category-sensitive retention, migration, and retrieval. Generic HSM products also implement policy-based demotion and access-triggered promotion. No retained source documents the entire proposed composition—quota-event audit, protected eligibility matrix, eligible-only LRU, distributional latency review, copy-preserving pilot, preregistered thresholds, and automatic rollback—as one deployed Earth-model repository intervention.","source_ids":["SRC3","SRC5","SRC6","SRC7","SRC8"],"uncertainty":"Operational architectures are documented, but comparative throughput, recall, duplicate-growth, group-incidence, and full lifecycle-cost results for the proposed pilot are absent."},"prior_art":{"disposition":"ESTABLISHED_PRACTICE","closest_analogues":[{"name":"CESM category-sensitive data-management and retention regime","source_ids":["SRC3","SRC4"],"same_problem":true,"same_causal_lever":false,"overlap":"It explicitly responds to finite storage for Earth-system output with differentiated treatment by data category, usage, anticipated demand, cost, and scientific value under named stewardship authority.","remaining_difference":"Its published plan emphasizes retention, compression, and selective deletion rather than a copy-preserving hot-to-cold pilot triggered by audited quota-blocked runs."},{"name":"ECMWF MARS/ECFS with disk caches and HPSS hierarchical storage","source_ids":["SRC6"],"same_problem":true,"same_causal_lever":true,"overlap":"Meteorological model data are routinely staged in disk caches, stored in a disk-and-tape hierarchy, retrieved through stable abstractions, and managed with associated metadata.","remaining_difference":"The public description does not report flat-quota counterfactuals, eligible-only LRU, equity metrics, an eight-week pilot, or comparison against adding 200 TiB of hot capacity."},{"name":"DKRZ climate-data disk-cache and automated tape archive","source_ids":["SRC5"],"same_problem":true,"same_causal_lever":true,"overlap":"A climate-specific center routinely channels tens of petabytes of annual simulation output through a relatively small disk cache into a much larger tape archive while supporting retrieval.","remaining_difference":"The page does not disclose classification rules, blocked-run evidence, recall distributions, full lifecycle costs, or rollback thresholds."},{"name":"IBM policy-based hierarchical storage management","source_ids":["SRC7"],"same_problem":false,"same_causal_lever":true,"overlap":"It supplies automatic policy-based migration from fast expensive media to cheaper media and dynamic retrieval when archived data are accessed.","remaining_difference":"It is domain-generic and does not supply Earth-science preservation governance, provenance safeguards, adopter incidence analysis, or the proposed causal audit."}],"contrastive_claim_remaining":"At one repository, after excluding unique, held, restricted, or uncertain objects and counting preservation, recall, stewardship, duplication, and access costs, at least 200 TiB can be reversibly demoted so that at least four runs previously blocked solely by hot quota are admitted, median quota wait falls by at least 48 hours, all safeguard thresholds hold, and the 90-day result is preferable to adding 200 TiB of hot capacity.","contrastive_claim_falsifier":"The audit finds fewer than 200 TiB eligible, cannot identify four runs blocked solely by hot quota or show that demotion would admit them, or finds that lifecycle cost is not at least $10,000 below expansion; alternatively, a passing-audit pilot fails either throughput outcome or breaches any recall, integrity, provenance, cost, duplication, or group-access safeguard.","confidence":"HIGH","search_limitations":"This bounded public-web search cannot inspect private quota-event logs, repository inventories, contracts, energy accounting, recall traces, user-group incidence, or pending procurements. It establishes mature analogous practice, not universal deployment or identity with every proposed safeguard."},"researchability_gates":{"externally_supported_problem":{"status":"PASS","rationale":"Primary research, an Earth-model management plan, and official quota documentation jointly establish that model output strains storage, can reduce scientific work, and can become unwritable at quota limits.","source_ids":["SRC1","SRC2","SRC3"]},"identifiable_adopter_or_authorizer":{"status":"PASS","rationale":"Modeling-program data managers, investigators, scientific steering bodies, repository operators, and institutional archive owners are externally identifiable decision makers.","source_ids":["SRC3","SRC5","SRC6"]},"distinct_testable_incremental_claim":{"status":"PASS","rationale":"Although tiering is established practice, the repository-specific causal claim remains distinct and falsifiable through the stated byte, blocked-run, admitted-run, wait-time, cost, recall, integrity, duplication, and access-incidence thresholds.","source_ids":["SRC2","SRC3","SRC5","SRC6"]},"bounded_next_evidence_step":{"status":"PASS","rationale":"A four-week read-only audit followed conditionally by one copy-preserving eight-week pilot is bounded, measurable, reversible, and directly tests the remaining local claim against hot-capacity expansion.","source_ids":["SRC2","SRC3","SRC6","SRC7"]},"no_unresolved_safety_or_authority_stop":{"status":"PASS","rationale":"Preservation and access are genuine constraints, but the proposed exclusions, preserved copies, integrity and provenance checks, governance approval, recall limits, and automatic promotion/rollback provide a credible path that does not require deletion. The initial audit itself does not alter stored objects.","source_ids":["SRC3","SRC6","SRC8"]},"adequate_search_evidence":{"status":"PASS","rationale":"The search covered all six required lanes, included older HSM terminology, official products and standards, German regional terminology, and component combinations. The eight retained sources span six independent publishers and include primary research, official guidance, a standard, and first-party operational systems.","source_ids":["SRC1","SRC2","SRC3","SRC4","SRC5","SRC6","SRC7","SRC8"]}},"strict_success":false,"screen_survival":false,"remaining_research_value":"MODERATE","recommended_next_step":"Seek the repository governance body's approval for a preregistered four-week audit using immutable quota-event, write-failure, occupancy, access, obligation, provenance, recall, duplication, staffing, energy, and cost records. Stop unless the exact audit gate passes; if it passes, run the copy-preserving eight-week pilot against the 200-TiB hot-expansion counterfactual with automatic rollback enabled.","world_novelty_boundary":"The search shows that hierarchical storage, disk-cache/tape architectures, differentiated Earth-model retention, migration, retrieval, and preservation governance are established practices. It does not establish that the exact audit-and-pilot composition has or has not appeared elsewhere, and cannot establish world novelty, patentability, freedom to operate, market size, or realized impact."}