Climate Weekly

Climate Weekly time series

Precipitation data: known miscalculation in legacy ETL

All precipitation variables (precip_min, precip_med, precip_max, precip_tot) used by this endpoint are affected by an ETL bug — the same described in the Climate endpoint. This directly impacts the precip_tot_sum field in the weekly output.

What was the problem?
The data source is the ERA5-Land reanalysis dataset, whose accumulation convention differs fundamentally from the standard ERA5 convention. In ERA5-Land, accumulations in short forecasts (steps 01 to 24) are accumulated from the beginning of the forecast to the end of the forecast step — i.e., from day=D, time=00:00 up to the step's valid time. At step 24, the timestamp day=D+1, 00:00 represents the total accumulation over the full 24 hours of day D. This means the time-stamped data at 00:00 actually represents the previous day's accumulation. The legacy ingestion pipeline incorrectly aggregated these accumulated variables using the ERA5 convention (where hourly steps represent the hour ending at the timestamp), producing shifted and incorrect daily precipitation totals that propagate into the weekly aggregation. See the ECMWF documentation for technical details.

What was fixed?
As of the latest update, the precipitation values have been recalculated using the correct ERA5-Land accumulation convention.

How to use the corrected data:
The precip_fixed query parameter defaults to true and controls all precip_* fields. When precip_fixed=true (default), the aggregation uses the corrected values. Set precip_fixed=false to retrieve the legacy (miscalculated) values for backward compatibility.

After 2026-08-01, the ingestion pipeline has been fixed. From that date onward, precip_fixed=true and precip_fixed=false return the same correct values — the toggle only affects historical data before the cutoff.

---

This endpoint is an aggregation of the Climate endpoint by Epiweek.

Parameters Table

Input

Parameter name	Required	Type	Description
*page	yes	int	Page to be displayed
*per_page	yes	int	How many items will be displayed per page (up to 300)
start	yes	int (YYYYWW)	Start epiweek
end	yes	int (YYYYWW)	End epiweek
geocode	no*	int	IBGE's municipality code
uf	no*	str (UF)	Two letters brazilian's state abbreviation. E.g: SP
macro_health_code	no*	int	5 digit brazilian's MacroHealth region geocode.
precip_fixed	no	bool	Use corrected precipitation values (default: true). See warning above.

Output (items)

Parameter name	Type	Description
epiweek	int (YYYYWW)	Epidemiological Week
geocodigo	int	IBGE's municipality code
temp_min_avg	float (°C)	Average minimum daily temperature
temp_med_avg	float (°C)	Average median daily temperature
temp_max_avg	float (°C)	Average maximum daily temperature
temp_amplit_avg	float (°C)	Average daily amplitude temperature
precip_tot_sum	float (mm)	Sum of total daily precipitation
umid_min	float (%)	Average minimum daily relative humidity
umid_med	float (%)	Average median daily relative humidity
umid_max	float (%)	Average maximum daily relative humidity

Details

One of the parameters is required: geocode, uf or macro_health_code page consists in the total amount of Items returned by the request divided by per_page. The pagination information is returned alongside with the returned request. E.g.:

'pagination': {
    'items': 10,                      # Amout of Items being displayed 
    'total_items': 10,          # Total amount of Items returned in the request
    'page': 1,                         # *request parameter
    'total_pages': 1,            # Total amount of pages returned in the request
    'per_page': 100             # *request parameter
},

Usage examples

PythonRcurl

import mosqlient

mosqlient.get_climate_weekly(
    api_key = api_key,
    start = "202201",
    end = "202301",
    # uf = "RJ",
    geocode = 3304557,
)

library(httr)
library(jsonlite)

climate_weekly_api <- "https://api.mosqlimate.org/api/datastore/climate/weekly/"

params <- list(
  page = 1,
  per_page = 300,
  start = YYYYWW,
  end = YYYYWW,
  geocode = MUNICIPALITY_GEOCODE,
  uf = UF,
  macro_health_code = MACROHEALTH_CODE
)

headers <- add_headers(
  `X-UID-Key` = API_KEY
)

resp <- GET(climate_weekly_api, query = params, headers)
items <- fromJSON(content(resp, "text", encoding = "UTF-8"))

curl -X 'GET' \
  'https://api.mosqlimate.org/api/datastore/climate/weekly/?start=YYYYWW&end=YYYYWW&page=1&per_page=300' \
  -H 'accept: application/json' \
  -H 'X-UID-Key: See X-UID-Key documentation'

# Or you can add a geocode and other filters
curl -X 'GET' \
  'https://api.mosqlimate.org/api/datastore/climate/weekly/?start=YYYYWW&end=YYYYWW&geocode=MUNICIPALITY_GEOCODE&uf=UF&macro_health_code=MACROHEALTH_CODE&page=1&per_page=300' \
  -H 'accept: application/json' \
  -H 'X-UID-Key: See X-UID-Key documentation'

*The response's pagination contain information about the amount of items returned by the API call. These information can be used to navigate between the queried data by changing the page parameter on the URL. See details