Testing the suitability of Marginal Distribution Sampling as a gap‐filling method using some meteorological data from seven sites in West Africa
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
Abstract Meteorological data are useful in many fields related to climate change studies and their use often requires them to be continuous. To date, marginal distribution sampling (MDS), which consists of filling a missing value with an average of the data that are found in similar meteorological conditions over a flexible time window, is widely adopted in the FLUXNET community. In this work, we evaluate the performance of MDS at diurnal and monthly scales for the incoming shortwave radiation (Swin), relative humidity (RH), vapour pressure deficit (VPD), air and soil temperatures (Tair, Tsoil) acquired across seven sites in West Africa. The criteria tested are the MDS's ability to (i) fill gaps while reducing the error rate, (ii) represent proper variability within data and finally (iii) ensure homogeneity between its output and original data. We found during the daytime that MDS is adequate for filling gaps in Swin when both reducing error rate and a good representation of variability are targeted. If the goal is to have a small error rate, then this approach is recommended for all investigated variables except VPD. During nighttime, MDS is satisfactory to minimize the error when filling gaps in Tair, Tsoil and RH while to represent their variabilities it becomes more sensitive to the rate of missing data. At a monthly scale, the gap‐filled data are consistent with the original ones for all variables attributable to data size and a wider sliding window that allows more data under similar conditions to be considered.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.003 | 0.002 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.002 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it