PaleoTech SA Climate Outlook Q1-2026

One comment

1. PaleoTech SA Climate Outlook Q1-2026 — Summary Review

   Narrative-basis review of the quarterly summary (ENSO · Wind · Rain · Temp) — judged on what each category claimed, not on deciles; sourced; specificity noted. Sub-regional and within-quarter variation is expected of a broad, sparse example outlook and not counted against it.

   Overall A weak quarter at approx 6-month lead — the outlook got the arc backwards. It leaned supportive/warm early, then weakening and more variable later; the quarter instead ran the opposite — a hot, dry January (−83% rainfall, fourth-warmest on record, 50 °C inland) flipping to an exceptionally wet, cooler February and March, second-wettest on record for both (Feb +356%, Mar +291%, wettest March since 1989). The two calls it got right were the ENSO state — a fading La Niña trending back to neutral by late Q1 — and TempMAP’s scorching January, with record inland heat and coastal moderation as flagged. Everything downstream of the rain arc missed. Rain’s “neutral-to-slightly supportive early for eastern & central SA” was doubly wrong: January was one of the driest on record and the only wet was the south-west, not the east or centre. And the record February–March wet that defined the quarter was called in the wrong direction and driven by inland/northern tropical moisture (Arkaroola 832% of normal), not the “southern frontal intrusions” the Wind module framed; the warm-lean held only for January, as the record rain cooled the days through February–March. The outlook even had the right idea that influence would shift to non-ENSO drivers late-quarter — it simply got the sign wrong, expecting drying where those drivers delivered record wet. Right ENSO state, wrong rain arc, blind to the record wet.

   PaleoTech SA Climate Outlook Q1-2026 — Module Reviews | Lead 6 months

   *Short, narrative, in-register module-by-module reviews — sourced; specificity noted. Internal indicators (proprietary precursors) are not deliberately exposed, so each module is judged on the state it predicted and its outcome, not its internals. (Reviews are against the published summary-level module calls; detailed Q1 module texts not supplied.)*

   1. ENSOLink™ — ENSO State Outlook
   Called: La Niña beginning to weaken through January–February, with a gradual return toward ENSO-neutral by late March. Indicators — tilt-anomaly signal weakening through late summer; Pacific-state energy redistribution trending toward neutral; subsurface cooling easing (reduced reinforcement of La Niña); cross-basin oscillation stabilising.
   Interpretation: a fading La Niña still provides some early-quarter moisture support, but predictability shifts back toward non-ENSO drivers (WindMAP, TempMAP, RainMAP) by late Q1.
   Happened: La Niña did weaken through Q1 and transitioned toward ENSO-neutral — NOAA forecast the shift for February–April 2026, still a weak La Niña early. And the late quarter was indeed run by non-ENSO drivers: the record-wet February–March came from inland/northern tropical moisture, not from ENSO. The one wrong note was “some early-quarter moisture support” — January was one of the driest on record (−83%), so the early support did not materialise.
   Held: strongly, and on two counts — the state call (fading La Niña → neutral by late Q1) ✅, and, notably, the structural call that predictability would shift to non-ENSO drivers by late Q1 ✅. ENSOLink correctly diagnosed both that ENSO was fading and that it would cede control — exactly what happened.
   Overcalled: only the token “early-quarter moisture support” — January was dry, not supported. Minor beside a correct fade and a correct handoff.
   Impact: crucially, ENSOLink did its job and passed the baton correctly — it flagged that late Q1 was no longer an ENSO story. The record-wet miss belongs downstream: the non-ENSO modules it handed to (WindMAP, RainMAP) then called those drivers as drying/variable when they delivered record wet. That is not ENSOLink’s error.
   Specificity & verdict: high specificity, verified on the state and on the handoff.
   Verdict: a strong module hit — arguably the best-structured call of the quarter — it knew ENSO was fading and correctly said the answer would come from non-ENSO drivers late-quarter. Only ding: the token “early moisture support.” The record-wet miss sits with the modules that took the baton, not with ENSOLink.

   2. WindMAP — Wind-State Outlook
   Called: strengthening westerly & south-westerly flow → variable mixed-flow by late March. Jan — La Niña pressure gradients; modest–strong south-easterly/easterly inland, enhanced westerlies along the Bight; short-burst wind events with troughs. Feb — La Niña weakens, subtropical ridge relaxes and repositions southward; winds shift west-southwest; more frontal intrusions into southern SA, especially coastal. Mar — localised wind surges in ridge–trough oscillations; ENSO to neutral, jet re-strengthens slightly in lower latitudes; transition to mixed-flow variability.
   Interpretation: an early-quarter window of supportive moisture transport, then increased volatility and reduced onshore coherence by late March.
   Happened: January was one of the driest on record (−83%) — no supportive moisture transport. February and March were record wet (+356% / +291%), but the wet was inland and northern (tropical moisture — Arkaroola 832% of normal), with the southern coast (Adelaide plains & ranges) a dry pocket in February — the opposite of the “frontal intrusions into southern SA, especially coastal” call.
   Held: little of the load-bearing content. “Increased volatility later” is loosely consistent with the event-driven wet back-half, and the March note that the “jet re-strengthens in lower latitudes” gestured at the right dynamical region (the subtropics/tropics) — but it drew the wrong conclusion from it (mixed-flow variability / reduced onshore coherence, not record wet).
   Calibration dings: the geography was inverted — it forecast a westerly/south-westerly/frontal regime feeding southern & coastal SA, when the wet came from inland/northern tropical incursions with the south dry. “Early-quarter supportive transport” failed (January dry). And it remains derivative of the ENSO-fade framing.
   Specificity & verdict: highly specific (month-by-month wind mechanics) and largely wrong where it counted — wrong regime (mid-latitude fronts vs tropical moisture), wrong geography (south/coast vs inland/north), wrong early timing. Verdict: a miss — this is exactly where ENSOLink’s correct handoff to non-ENSO drivers was fumbled. Its one flicker of insight (the lower-latitude jet) points at the fix: a tropical-moisture read, which WindMAP framed as variability rather than the record wet it produced.

   3. TempMAP — Temperature Outlook
   Called: Jan — slightly cooler than expected early (fading La Niña, stronger cloud-cover effects in places). Feb — near-average, with increased small heat spikes aligned to WindMAP volatility. Mar — above-average warmth, especially inland SA, with heat pulses in late-month transitions.
   Interpretation: begins mild, trends average mid-quarter, ends warmer with more rapid swings as ENSO decouples — a warming arc through the quarter.
   Happened: January was exceptionally HOT — mean +3.35 °C (4th-warmest on record), max +4.21 °C (3rd-warmest), 50.0 °C at Andamooka and Port Augusta — the opposite of “slightly cooler early.” February near-average overall (cool days, warm nights) — roughly matches “near-average.” March was COOL — max −1.23 °C (lowest March max since 2011), mean −0.17 °C — the opposite of “above-average warmth, especially inland.”
   Held: only February (near-average). The arc was backwards — TempMAP forecast warming through the quarter (mild → average → warm); reality cooled through it (record-hot January → cool, wet March).
   Miss + correction: The module’s own detailed January call was “slightly cooler than expected” — the opposite of the 4th-warmest January on record — so on its detail TempMAP missed the hot January, and the outlook is internally inconsistent (summary “warm/inland heat” vs detail “cooler early”). March was also missed (called above-average inland; was the coolest March maxima since 2011, cooled by the record rain).
   Specificity & verdict: highly specific and largely wrong — the temperature arc ran backwards, and the biggest, most falsifiable event (a 4th-warmest January) was called cool. Verdict: a miss — and a correction to the earlier over-credit. TempMAP was not Q1’s saving grace; only its February near-average call held. Even the strongest Q4 module missed Q1 badly once the wet/tropical regime took over — and its detailed January call was simply wrong.

   4. RainMAP™ — Rainfall Outlook
   Called: Jan — slightly above-average rainfall probability; moisture transport still partly La Niña-driven; localised storm-cell potential. Feb — neutral to slightly positive, strong regional variation. Mar — declining rainfall support as ENSO fades; increased dry-air intrusions; higher risk of late-summer dry spells.
   Interpretation: starts supported (especially January), becomes regionally split in February, then weakens — most rain early, increasingly patchy as Q1 concludes.
   Happened: January −83% — one of the driest on record (only the south-west wet), the opposite of “slightly above-average.” February +356% — second-wettest on record (inland/north huge, Adelaide a dry pocket). March +291% — second-wettest, wettest since 1989, eastern records — the opposite of “declining support / late-summer dry spells.”
   Held: only the fragment “strong regional variation” (February did split — inland drenched, Adelaide dry). Everything directional was backwards.
   Miss: the arc was inverted at both ends and at record magnitude. It called January supportive (was the driest) and explicitly forecast March “declining support” and “late-summer dry spells” in what became the wettest March since 1989. RainMAP forecast a declining quarter (wet early → dry late); reality was a record-increasing one (driest early → record-wet late).
   Note (summary vs detail): the summary was directionally better than the detail on the late quarter — it said “more variable and event-driven later … higher storm-driven variability inland,” which leans toward the late inland event-driven wet, whereas the detail said the opposite (“declining support … late-summer dry spells”). Both missed January. So my earlier summary-level credit for a “late-inland-event-driven” lean was fair to the summary; it is the detail that reversed it.
   Specificity & verdict: highly specific and comprehensively wrong on direction — the starkest reversal of the four, forecasting late-summer dry spells into a record-wet March.
   Verdict: a clear miss — no read on the tropical-moisture regime that delivered the record wet, defaulting to the “La Niña fading → drying” prior at exactly the wrong time.

   Q1 2026 — module scorecard (synthesis)
   MODULE VERDICT INDEPENDENT VALUE
   🌊 ENSO (ENSOLink) ✅ state held (fading → neutral) + correctly handed off to non-ENSO drivers high — right call and right limits; only the token “early support” missed
   💨 Wind (WindPulse) ✖️ miss — wrong timing, geography, mechanism low — ENSO restated, inherited the wrong prior
   🌧️ Rain (RainMAP) ✖️ arc backwards at both ends — forecast late-summer dry spells into the wettest March since 1989; only Feb “regional variation” held low — no read on the tropical wet
   🌡️ Temp (TempMAP) ✖️ miss — arc backwards; called cool Jan (was 4th-warmest), warm Mar (was coolest since 2011); only Feb near-avg held low — regime broke it; summary/detail inconsistent on Jan

   Cross-cutting patterns:
   The quarter’s defining event — record wet February–March — was missed, and on the detailed module texts the miss is broad: only ENSOLink comes out ahead. ENSOLink did its job — it called the fade to neutral and correctly said predictability would shift to non-ENSO drivers by late Q1. The modules that took the baton then failed: WindMAP called southern/coastal fronts (the wet fell inland/north), RainMAP called the arc backwards, and TempMAP — the Q4 star — got the temperature arc backwards too (its detail called January cool when it was the 4th-warmest on record, and March warm when it was the coolest since 2011). The failure mode is a regime the non-ENSO modules don’t track: inland/northern tropical moisture (a northwest-cloudband-type pattern) that overrode the fading La Niña; built for ENSO/ridge dynamics, they defaulted to “La Niña fading → drying/warming-late,” exactly the wrong prior. Where Q4’s misses were of magnitude, Q1’s were of direction — arc, geography and mechanism all inverted. One strong call this quarter: ENSOLink’s fade-and-handoff. Note also an internal inconsistency to reconcile — the TempMAP summary (“warm/inland heat”) and its detail (“cooler early”) disagree on January, and the accepted summary-review “Overall” line credits “TempMAP’s scorching January,” which the module’s own detail undercuts. Two quarters in: credible when ENSO/ridge dynamics lead (Q4), broadly exposed when tropical/non-ENSO moisture leads (Q1) — the standing gap is a tropical-moisture signal for the non-ENSO modules to read once ENSOLink hands off.

   Root cause — ENSO over-weighted. Beneath all of the above sits one diagnosis: the outlook weights ENSO too heavily for a region where its signal is weak. Every Q1 module reasoned from the fading La Niña (“La Niña-aligned,” “as ENSO fades → declining,” “cooler early due to fading La Niña”), so a single ENSO-derived prior propagated the same directional error through Wind, Rain and Temp — one error, four modules. SA’s ENSO rainfall signal is modest (ONI ~−0.16 in the analog work; the ridge/STR and other drivers dominate), and a −0.5 °C fading La Niña is weaker still — yet it anchored the quarter. ENSOLink correctly said to hand off to non-ENSO drivers, but nothing was built to catch the baton, so the modules fell back on the ENSO prior by default. ENSO is also slow and smooth, so weighting it heavily biases the outlook toward persistence and blinds it to sub-seasonal regime flips — exactly the tropical-moisture flip that ran Q1. Fix: weight ENSO by strength × SA relevance (near-zero when weak/fading) and give the non-ENSO drivers — tropical-moisture / NW-cloudband, STR, SAM — real independent weight so they can override when ENSO cedes control. (Inferred from the module reasoning and the shared failure mode; internal weights are not exposed.)

   Process finding — the summary beat the detail. For both TempMAP and RainMAP the SUMMARY was directionally better than the DETAILED monthly text. TempMAP’s summary (“warm-leaning, inland heat spikes”) aligned with the record-hot January; its detail called January cool. RainMAP’s summary (“more event-driven, higher storm-driven variability inland later”) leaned toward the late inland wet; its detail called “late-summer dry spells.” In both, the detailed monthly arc added confident error, not skill — the extra specificity committed to the wrong sequence while the coarser summary stayed aligned. The falsifiable calls live in the detail, but here that is also where the direction went wrong: the modules resolve a quarter-level lean better than a month-by-month arc, and over-committing to the arc cost skill.

   Sources: ENSO — WMO Feb 2026; NOAA CPC Mar 2026; transition to neutral Feb–Apr 2026 (still weak La Niña early). Rain & Temp — BoM South Australia in January / February / March 2026.