↖ CPA Weather Lab
PRISM SLAM · Integrated real-data studio

Teach → Experiment → Regionalize

One self-contained laboratory joining the step-by-step SLAM math, exact L-moment estimator, completed 1981–2025 PRISM experiment, real 5×5 CAM fields, annual winning storms, a kurtosis microscope, and an explicit bridge into regionalization. It does not pretend the 10,000-sample target bootstrap has already been run.

45Jun–Sep seasons
2,376accepted 5×5 boxes
106,920annual box winners
2,673,000winning event cells
59,400CAM values
PASScanonical analysis
OrientationTeach the mathReal experimentBox laboratory Kurtosis microscopeCalculatorRegionalizationCheck intuition
Orientation

Keep these three objects separate

1 · Annual winning event

For each box and year, test every contained 72-hour period and choose the period with the greatest cell-area-weighted 5×5 mean. Freeze all 25 cell totals from that same event.

2 · CAM field

At each fixed relative cell position, average that cell across the 45 selected annual events. The result is one 5×5 field with 25 CAM values.

3 · Spatial L-moments

Sort those 25 CAM values and compute L-mean, L-CV, L-skewness and L-kurtosis. The observations at this stage are cells in space, not years in time.

Two different averages are intentionally used. Event selection uses the spherical cell-area-weighted 5×5 mean. Once the 25 CAM values exist, L₁ is their ordinary unweighted arithmetic mean. Do not merge those operations.
Daily PRISMJun–Sep
72-hour totalscontained
Annual winnerarea-weighted
25-cell eventfrozen together
45 winners1981–2025
25-cell CAMcell-wise mean
4 L-momentsspatial shape
Regionalizelater inference
Teach · from first principles

From a rainfall field to L₁, L₂, L₃ and L₄

This integrates the useful pieces of the earlier teaching pages into one chain rather than making you mentally teleport among three files.

A small synthetic CAM

After annual event selection and cell-by-cell averaging, suppose a five-cell CAM is:

22.5, 30, 35, 45, 60 mm

Sort it:

22.5 ≤ 30 ≤ 35 ≤ 45 ≤ 60
The real experiment uses 25 CAM cells. Five cells are used here only so every rank weight fits on the screen.

The rank-weighted averages

br = (1/n) Σ [ C(i−1,r) / C(n−1,r) ] x(i)

b₀ weights every rank equally. As r rises, higher ranks receive progressively greater weight. These are building blocks, not new rainfall observations.

b₀
38.5
ordinary mean
b₁
23.75
upper ranks emphasized
b₂
17.6667
stronger upper-rank emphasis
b₃
14.25
strongest upper-rank emphasis

Dimensional L-moments

L₁ = b₀
L₂ = 2b₁ − b₀
L₃ = 6b₂ − 6b₁ + b₀
L₄ = 20b₃ − 30b₂ + 12b₁ − b₀

The four SLAM descriptors

L-mean = L₁
L-CV = τ₂ = L₂/L₁
L-skewness = τ₃ = L₃/L₂
L-kurtosis = τ₄ = L₄/L₂
Why L-moments instead of ordinary moments? Ordinary variance/skewness/kurtosis use squared, cubed and fourth-power deviations. These L-moments are linear combinations of ranked observations, so one spectacular cell has less power to hijack the statistic.
Experiment · completed PRISM run

See the real domain before interpreting any statistic

PRISM experiment domain
11,220 geometric target cells, 10,421 valid cells, 799 persistent no-data cells, and grid envelopes.
5x5 moving window coverage
How many accepted 5×5 windows contain each valid PRISM cell.
Valid PRISM cells
10,421
Persistent no-data
799
Accepted windows
2,376
Rejected 5×5 windows
197

Interactive metric map: all 2,376 real boxes

Move over a point for exact values. Click a representative box to load it into the Box Laboratory.
L-mean median
72.11 mm
55.60 to 88.80
L-CV median
0.0160
0.0031 to 0.0701
L-skew median
-0.044
-0.478 to 0.448
L-kurt median
0.080
-0.099 to 0.420

A real geographic signal

Across the 2,376 boxes, longitude and L-mean correlate at r ≈ +0.860. Eastward boxes are substantially wetter in the 45-year CAM magnitude. The other L-moment ratios do not simply reproduce that gradient.

Approx. longitude thirdnL-meanL-CVL-skewL-kurt
West third80062.580.0177-0.0310.083
Central third81071.580.0187-0.0240.089
East third76680.700.0163-0.0730.089
These thirds are descriptive bins for learning, not official climate regions and not the final SLAM regionalization.

Why this matters for regionalization

A regionalization based only on magnitude would mostly rediscover the east–west precipitation gradient. SLAM asks for agreement in magnitude, relative spread, asymmetry and tail structure. The extra dimensions are what prevent “same average rainfall” from being mistaken for “same spatial storm texture.”

Experiment · actual storms and actual CAMs

Box Laboratory: watch 45 annual winners collapse into one CAM

Selected annual 72-hour winner

45-year CAM for the same 25 relative cells

Each CAM cell is the mean of that fixed relative cell across the 45 annual winning events.

Observed CAM L-moments

One bootstrap CAM, for intuition only

Click the amber button. It resamples the 45 annual event fields with replacement, preserves the 25-cell field within each selected year, rebuilds one CAM, and recomputes the four L-moments. This is a classroom demonstration, not the 10,000-replicate regionalization run.
Deep dive · the hard one

L-kurtosis is not “how pointy the map looks”

Best working intuition: after accounting for overall spatial spread with L₂, L-kurtosis asks whether that spread is concentrated unusually far out in the outer ranks of the 25-cell CAM distribution, rather than being distributed smoothly through the shoulder ranks.

For n = 25, L₄ has a rank signature

Because L₄ is a linear combination of b₀…b₃, each sorted CAM value gets one net L₄ coefficient. Center the values by subtracting L₁ and the pattern becomes interpretable:

positive net L₄ coefficientnegative net L₄ coefficient

The coefficient sum is zero, so shifting every cell by +50 mm does not change L₄. The dry extreme has a negative coefficient and is below the mean, so their product contributes positively; the wet extreme has a positive coefficient and is above the mean, so it also contributes positively. Shoulder ranks tend to contribute in the opposite direction. Scaling all values scales L₄ and L₂ together, leaving τ₄ unchanged.

The ratio that SLAM uses

τ₄ = L₄ / L₂

L₄ is in millimeters. L₂ is the spatial L-scale in millimeters. Dividing makes τ₄ dimensionless: tail/shoulder structure relative to the field's overall spread.

High τ₄ does not mean “the box has huge spread.” A box can have ordinary L-CV but high L-kurtosis if most cells cluster tightly while a few outer-ranked cells sit distinctly away from that core.

Kurtosis example

Sorted CAM values and L₄ contributions

Actual low-L-kurtosis box

B00827_R0239_C1119 · τ₄ = −0.0988. Its 25 CAM values progress fairly smoothly from about 60.0 to 65.9 mm. The shoulder ranks carry much of the spread, so the outermost ranks do not dominate after normalization.

Actual high-L-kurtosis box

B02180_R0205_C1173 · τ₄ = +0.4203. Most CAM cells cluster roughly around 72.8–75.1 mm, while the far upper ranks jump to about 79.5 and 81.9 mm and the low end reaches about 69.6–70.1 mm. That separation of outer ranks from the shoulders drives positive L₄.

Do not import the ordinary-kurtosis cliché too literally. “Peakedness” is an unreliable mental shortcut here. These are 25 spatial CAM values and τ₄ is built from ranked linear weights, not fourth powers. Think outer ranks versus shoulders, normalized by L-scale. A positive τ₄ means the outer ranks are more separated from the middle/shoulder ranks than the L₂ scale alone would suggest; a low or negative τ₄ means the spread is carried more continuously through the shoulders rather than concentrated at the extremes.
Experiment · exact estimator

Paste any CAM values and expose every calculation

This combines the earlier estimator calculator with the real-box laboratory. Load a representative 25-cell CAM or type your own values.

Probability-weighted moments

Convert to L₁–L₄

Rank weights and contributions

ix(i)w₁xw₁w₂xw₂w₃xw₃
Regionalization · concept + real pre-bootstrap sandbox

What “regionalization” will mean in this PRISM adaptation

Target uncertainty

Select the exact target/reference box. Resample its 45 complete annual 25-cell event fields with replacement. Rebuild one CAM per replicate and compute four L-moment statistics. Repeat 10,000 times with a fixed seed.

Translated comparisons

Every accepted translated 5×5 box already has its observed L-mean, L-CV, L-skewness and L-kurtosis. Compare each observed statistic with the corresponding target bootstrap distribution to obtain two-sided empirical p-values.

Spatial domain

Control false discoveries for each statistic, combine the four acceptance fields, then keep the connected accepted region associated with the target. That connected set is the inferential transposition region.

Current status: we have not selected the final target and have not run the 10,000-replicate bootstrap. The sandbox below is deliberately not a p-value map. It is an intuition tool showing how multivariate similarity can shrink or reshape a region.

Pre-bootstrap similarity sandbox

What this sandbox teaches correctly

Regionalization is an intersection problem. A box can look similar in mean rainfall and still fail because its within-box contrast, asymmetry or tail structure differs. Adding dimensions usually reduces the candidate set.

What this sandbox does not claim

Standard-deviation tolerances are not SLAM p-values, do not represent target sampling uncertainty, do not apply FDR, and do not enforce the final connected-component rule. This exists to build intuition before the inferential machinery is turned on.

Check intuition

Four quick tests before bootstrap

No. L-mean describes overall CAM magnitude; L-CV describes spatial spread relative to that magnitude.
No. At the L-moment stage the observations are the 25 CAM cells. Positive L-skewness means the upper side of the spatial CAM distribution is more extended than the lower side.
Not necessarily. τ₄ is normalized by L₂. High τ₄ means the outer ranks are unusually separated relative to the shoulder ranks and overall L-scale.
No. Resample whole annual winning event fields. The 25 cells within an event must remain together so the observed spatial dependence is preserved.
← CPA Weather Lab · Learn