Possible venting date ranges for Brazil

[benshostak]

Added file "Brazil_Venting_Ranges_20150627_20250801.csv"

[jonaraphael]

@benshostak , can you clarify what changed and why between these two documents? No settings, configuration, or code changed between the commits, so it's really unclear how the output could possibly be different...

[benshostak]

@jonaraphael The changes I made to the code were on a different branch (offshore flaring) that I’m having trouble committing, so aren’t being reflected in the “Brazil-Possible-Venting-from-Flaring-Sites” branch. I believe I am running into the same git issue I had before when committing with the “ruff” package but forgot how to reconcile it :(

I added this to the code prior to processing the points from skytruth annual flaring dataset:

function to filter out rows where flaring only occurred prior to s2 imagery

def filter_recent_flaring(fc):
bcm_fields = [
“bcm_2016”,
“bcm_2017",
“bcm_2018”,
“bcm_2019",
“bcm_2020”,
“bcm_2021",
“bcm_2022”,
“bcm_2023",
]

# Start with first field
combined_filter = ee.Filter.gt(bcm_fields[0], 0)

# Combine the rest using ee.Filter.or
for field in bcm_fields[1:]:
    combined_filter = ee.Filter.Or(combined_filter, ee.Filter.gt(field, 0))

# Apply filter to FeatureCollection
return fc.filter(combined_filter)

points_filtered = filter_recent_flaring(table)
print(“points 2016 - 2023:“, points_filtered.size().getInfo())

Ethan pointed out some discrepancies between the gfw infra dataset that he reviewed and the flaring/venting dataset. Turns out around 40+ points in the flaring dataset were from before S2 imagery became available (2012 - 2015) and the infrastructure was not visible in the S2 imagery from 2016 onwards. I added the code above to remove points where no flaring was detected between 2016 and 2023 to remove these points.