html`

This notebook matches radionuclide measurements in marine biota with contemporaneous seawater measurements, based on spatial and temporal proximity, to derive concentration ratios (CR = activity in biota / activity in seawater, both in Bq/kg wet weight) for matched pairs.

Datasets

The biota dataset is the curated dataset prepared for the LaMer CRP seafood dose assessment, exported from MARIS: https://maris.iaea.org/home/news/2145. The seawater dataset comprises all available measurements of the same 16 radionuclides exported from the MARIS database in May 2026.

Matching

For each biota record, seawater measurements within a user-defined radius and time window are identified using a three-stage pre-filter followed by precise distance calculation. First, a binary search on sorted latitude narrows candidates to those within the latitudinal extent of the search radius. Second, a longitude range check (adjusted for latitude) removes records outside the longitudinal extent. Third, a day-difference filter removes records outside the time window. Precise distance is then computed for surviving candidates only using the Haversine formula (via turf.js), keeping only those within the specified radius.

Mean seawater activity and propagated uncertainty are computed across all matching records.

Quality Score

A quality score (0–1) combines normalised spatial and temporal proximity using a weighted geometric mean, and can be used to threshold accepted matches. For each matched seawater record i, the quality score is computed as a weighted geometric mean of normalised spatial and temporal proximity:

${tex.block`q_i = \left(1 - \frac{d_i}{r}\right)^{w_s} \cdot \left(1 - \frac{\Delta t_i}{\Delta T + 1}\right)^{w_t}`}

where dᵢ is the distance (km), r is the maximum radius (km), Δtᵢ is the absolute difference in sampling dates (days), ΔT is the maximum time window (days), and w_s and w_t = 1 − w_s are the spatial and temporal weights. The +1 in the denominator ensures boundary matches score just above zero rather than exactly zero.

The reported quality score Q is the arithmetic mean of qᵢ across all N matched seawater records:

${tex.block`Q = \frac{1}{N}\sum_{i=1}^{N} q_i`}

Q ranges from 0 (poor match) to 1 (perfect co-location and co-timing).

Units

Seawater values reported in Bq/m³ are converted to Bq/kg assuming a seawater density of 1.025 kg/L. Measurements in incompatible units (e.g. DELTA/mill, kg/kg) are excluded.

Statistics

CR distributions are assumed lognormal; geometric mean and geometric standard deviation (GSD) are reported alongside arithmetic statistics. Use the controls below to select nuclide, biogroup, matching radius, time window, spatial/temporal weighting, and minimum quality threshold.

To do

Check the consistency of seawater data - filtered or not
Depth filter
Decay correct seawater values to the sampling date of the biota measurement (for completeness)

Controls