Literature
The evidence base
Relevance scores are produced by 02_local_pdf_miner.py over the PDFs in the project folder — they count matched variables, sensors, methods and places, so they are measurements rather than editorial judgement. Every DOI links to the publisher.
8
In the folder
mined and catalogued
4
Cited but missing
referenced in the config, PDF absent
9
Core relevance ≥ 5
of 6 possible
14
Paper → variable links
which paper justifies which variable
In the folder
| Paper | Venue | DOI | Rel. | What it contains | How it is used | Variables |
|---|---|---|---|---|---|---|
| Jutla, Akanda & Islam 2010 | J. Am. Water Resour. Assoc. | 10.1111/j.1752-1688.2010.00448.x ↗
|
6/6 | Coastal chlorophyll to cholera tracking with SeaWiFS and AVHRR over the Bay of Bengal. | The origin of the coastal-intrusion hypothesis that this study's gradient design tests. | CHL_OCX |
| Nusrat et al. 2024 | GeoHealth | 10.1029/2024GH001039 ↗
|
6/6 | The Matlab baseline: a 1 km monthly weighted-sum cholera risk index over a 22x22 km domain that matches your shapefile. | Benchmark M4 against it. Produce an AOI-aggregate series so the comparison is direct. | CHL_OCX |
| Anas et al. 2021 | Remote Sensing 13(5):1034 | 10.3390/rs13051034 ↗
|
5/6 | V. cholerae dynamics in a tropical lake and estuary using Sentinel-2 MSI, Sentinel-3 OLCI and Landsat. | The closest methodological template to what you are building. Follow its ACOLITE workflow. | FAI MCI NDCI PHYCOCYANIN |
| Jutla et al. 2013 | Environmental Modelling & Software | 10.1016/j.envsoft.2013.05.008 ↗
|
5/6 | Prediction framework introducing the Himalayan winter air-temperature discharge proxy. | Source of the SWM variable and the 6-11 month lead structure. | CHL_OCX SWM |
| Lobitz et al. 2000 | PNAS | no DOI in the PDF | 5/6 | The founding paper - SST and SSH as indirect V. cholerae detection. | Cite for mechanism, not method. The sensors are long superseded. | CHL_OCX |
| Env. Econ. & Management 2024 | J. Environ. Econ. Manage. 125:102966 | 10.1016/j.jeem.2024.102966 ↗
|
4/6 | Salinity, flooding and health economics in Bangladesh. | Context for the salinity mechanism and WASH covariates. Not a remote-sensing method source. | |
| Jutla et al. 2012 | Remote Sensing of Environment | 10.1016/j.rse.2012.03.005 ↗
|
4/6 | Space-time plankton variability in the Bay of Bengal, measuring the autocorrelation structure. | Source of the finding that daily 9 km chlorophyll has lag-1 autocorrelation of 0.20 - it is white noise. This is why ocean colour stays monthly. | CHL_OCX |
| Marine litter, Dniester 2025
NOT USED |
Marine Pollution Bulletin | 10.1016/j.marpolbul.2025.117602 ↗
|
1/6 | Numerical modelling of marine litter dispersion. No cholera, no Bangladesh. | NOT USED. Scored 1/6 on relevance - remove it from the corpus. |
Titles and relevance scores read directly from 02_Downloads/pdf_mined_summary.csv.
Cited in the config but missing from the folder
These four are referenced by the pipeline and are not present. Three of them define variables that
are already being computed, which means the code currently implements algorithms whose primary sources
are not in the corpus. Add them before writing the methods section.
| Paper | Venue | DOI | Why it matters | Consequence of it being absent | Variables |
|---|---|---|---|---|---|
| Hu 2009 MISSING |
Remote Sens. Environ. 113(10):2118-2129 | 10.1016/j.rse.2009.05.012 ↗
|
The Floating Algae Index. | MISSING FROM THE FOLDER. Add it - FAI is a CORE variable at all three sites. | FAI |
| Islam et al. 2007 MISSING |
J. Appl. Sci. 7(13):1785-1789 | — | The cyanobacteria reservoir mechanism - the theoretical basis for phycocyanin as a cholera predictor. | MISSING FROM THE FOLDER. It is cited throughout aoi_config.json but the PDF is not in Paper/. Phycocyanin is the highest-novelty variable in this study and its key citation is absent. | PHYCOCYANIN |
| Mishra & Mishra 2012 MISSING |
Remote Sens. Environ. 117:394-406 | 10.1016/j.rse.2011.10.016 ↗
|
The NDCI algorithm and its chlorophyll polynomial. | MISSING FROM THE FOLDER. Add it - NDCI is your primary inland chlorophyll variable. | NDCI |
| Sarker et al. 2019 MISSING |
Front. Mar. Sci. 6:598 | — | Locally fitted chlorophyll retrieval coefficients for Bangladeshi water. | MISSING FROM THE FOLDER. The cheapest route to turning your ordinal indices into defensible concentrations. | NDCI |
How to fetch them
The harvester in the project folder searches five scholarly APIs and pulls open-access PDFs through Unpaywall:
cd 01_Scripts
python 01_harvest_literature.py
# then re-mine the folder so this page updates
python 02_local_pdf_miner.py