Moncton data projectHow many days did Moncton have snow on the ground in 2025? Reading the French file
Environment and Climate Change Canada publishes daily weather records for the Moncton airport station, in French and in English. In our classes the older children pick the French file deliberately, since it teaches two things the English file hides. Every count here is ours.
Lesson 1: the decimal comma. The French file writes 3.1 as 3,1, and Python refuses to read that as a number. Our three-line test shows the error and the fix:try:
print(float("3,1"))
except ValueError as error:
print("ValueError:", error)
print(float("3,1".replace(",", ".")))It prints ValueError: could not convert string to float: '3,1' and then 3.1. The column names are French too: Mois is month, Temp max.(°C) is the daily high and Neige au sol (cm) is snow on the ground.
Days with snow on the ground at the Moncton airport station in 2025, by month: our count from ECCC daily data (station MONCTON/GREATER MONCTON ROMEO LEBLANC INTL A, Climate ID 8103201), French file downloaded 22 October 2026| Month | Days with snow on the ground | Days with no value recorded |
|---|
| January | 25 | 0 |
| February | 28 | 0 |
| March | 19 | 2 |
| April | 4 | 23 |
| May to October | 0 | every day |
| November | 3 | 22 |
| December | 23 | 0 |
Our script reads the French file directly. Lesson 2: a blank is not a zero. When the snow column is empty, the script counts the day as not recorded and moves on, instead of pretending there was no snow:import csv
snow_days = {}
missing = {}
with open("fr_climate_daily_moncton_50309_2025.csv", encoding="utf-8-sig") as f:
lignes = csv.DictReader(f)
for row in lignes:
mois = int(row["Mois"])
text = row["Neige au sol (cm)"]
if text == "":
missing[mois] = missing.get(mois, 0) + 1
continue
depth = float(text.replace(",", "."))
if depth > 0:
snow_days[mois] = snow_days.get(mois, 0) + 1
for mois in range(1, 13):
print(mois, snow_days.get(mois, 0), "days with snow on the ground,", missing.get(mois, 0), "days not recorded")
February had snow every day
All 28 February days show snow on the ground. January had 25 and December 23. Across the year we count 102 days with snow on the ground, a figure we counted directly rather than adding up the months.
Why April is uncertain
April shows only 4 snow days, but 23 April days have no value at all. So the honest statement is that at least 4 April days had snow on the ground; the file cannot tell us about the rest.
Two files, one record
We checked the French and English downloads against each other: same 365 dates, same 31 columns in the same order, and all 365 daily highs agree once the comma is swapped for a point.
What the file cannot tell you
One airport station is not the whole city, and the snow column is simply left empty for most of the warm months. A careful student says the station recorded snow on the ground on 102 days in 2025, not that Moncton had exactly 102 snowy days.
Data Source: Environment and Climate Change Canada, historical climate data, used under the Environment and Climate Change Canada Data Services End-use Licence. This is federal data about Moncton rather than a City of Moncton file, which we chose because its licence is perpetual and allows commercial reuse. Neither the data nor its licence implies any endorsement of our classes by ECCC.
