Algae Dataset
Every data hub, dataset, sensor, variable, paper and limitation in one tree — and for each: where it comes from, how far back it goes, how often it captures, why it belongs in a cholera study, and why it does not work everywhere.
- +DATA HUBSwhere the data physically comes from
- +DATASET TYPESgrouped by what kind of water they see
- +SITE-WISEwhat actually applies at each site
- +DURATION & CAPTURE CADENCEdaily · weekly · 10-day · 16-day · monthly
- +VARIABLESwhat is computed from the bands
- −LIMITATIONSwhy not — typed and severity-ranked
- +VALIDATION2 limitations · 1 critical
- +ACCESS6 limitations · 4 critical
- +SPATIAL6 limitations · 3 critical
- +TEMPORAL7 limitations · 2 critical
- +ALGORITHMIC3 limitations · 1 critical
- −COVERAGE6 limitations · 1 critical
- ·Dhaka has almost no pre-2015 algae observationCRITICALsite level
- ·Copernicus LWQ300 lake mask may exclude all three sitesHIGHdataset AL09
- ·Landsat 7 SLC-off stripes after May 2003HIGHdataset AL05
- ·IRI is a strong climate archive but a thin ocean-colour o…HIGHdataset AL17
- ·Multiple testing across the lag gridMEDIUMstudy level
- ·Dhaka polygon is a GADM level-3 unit labelled TejgaonMEDIUMsite level
- +ATMOSPHERIC1 limitations
- +SPECTRAL2 limitations
- +PAPERSthe evidence, DOI-linked
COVERAGE limitations
CRITICAL Dhaka has almost no pre-2015 algae observation coverage blocks Dhaka
Every Dhaka water body is below 250 m, so MODIS and Sentinel-3 are both unusable. Landsat at 16 days is the only option before 2015, giving roughly 5 percent of days observed - a daily series for 2000-2014 would be about 95 percent invented.
HIGH Copernicus LWQ300 lake mask may exclude all three sites coverage
The Lake Water Quality product only reports for water bodies in its predefined mask at 300 m. Coastal lagoons are often excluded by design.
HIGH Landsat 7 SLC-off stripes after May 2003 coverage
The scan-line corrector failed in 2003, leaving wedge-shaped data gaps that widen toward the scene edge and lose roughly 22 percent of each image.
HIGH IRI is a strong climate archive but a thin ocean-colour one coverage
The Ingrid engine is excellent for SST, rainfall and climate indices, with some records reaching the 1850s. Its NASA/PO-DAAC/oceancolor branch is sparse and inconsistently maintained.
MEDIUM Multiple testing across the lag grid coverage
34 variables x 14 lag steps x 3 sites is 1,428 cross-correlations. At p<0.05 roughly 70 will clear significance by chance alone, and the strongest of those will look publishable.
MEDIUM Dhaka polygon is a GADM level-3 unit labelled Tejgaon coverage blocks Dhaka
Dhaka_City.shp has NAME_3 = 'Tejgaon' but an extent of 17.6 x 26.0 km, which is city-scale rather than thana-scale. It works fine as a city AOI, but the label is misleading.