Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/maxiricalde/ProfeLedesma/llms.txt

Use this file to discover all available pages before exploring further.

Overview

This workshop is built around a real-world solar irradiance dataset spanning four years (2020–2023) of 1-minute Global Horizontal Irradiance (GHI) measurements from station LQ, located in the La Quiaca area of northwestern Argentina at an altitude of 3,500 m. The four annual CSV files together contain approximately 1,764,898 raw 1-minute readings before any deduplication or resampling (2020: 526,994 rows; 2021: 503,874; 2022: 492,567; 2023 partial: 241,463). Alongside the measured time series, the dataset is paired with a modeled clear-sky reference generated by the ARGP clear-sky model. The modeled data covers a longer historical window, starting from 2009, and serves as a physically consistent benchmark for quality-control, gap-filling, and validation throughout the preprocessing pipeline.

Measured Data — measured/GHI_LQyyyy.csv

The raw pyranometer files are stored as one CSV per year, named GHI_LQ2020.csv through GHI_LQ2023.csv. Each file contains the following columns:
ColumnDescription
AÑOCalendar year of the measurement
DIA JULIANOJulian day of year (1–365/366)
HORAMinute-of-day counter (1-based)
IRRADIANCIA (W/m2)Measured GHI in watts per square metre
FechaISO-8601 timestamp (YYYY-MM-DD HH:MM:SS)
Sample rows from GHI_LQ2020.csv:
AÑO,DIA JULIANO,HORA,IRRADIANCIA (W/m2),Fecha
2020,1,1,-3.12837109,2020-01-01 00:01:00
2020,1,2,-3.236245955,2020-01-01 00:02:00
2020,1,3,-3.236245955,2020-01-01 00:03:00
Note: Negative values near zero are normal and represent instrument noise during the night-time baseline.
The two columns actually used when loading the data are column index 4 (Fecha) and column index 3 (IRRADIANCIA (W/m2)).

Modeled Data — modelated/LQ.csv

The ARGP clear-sky model output is stored in modelated/LQ.csv and contains two columns:
ColumnDescription
dateISO-8601 timestamp
GHIModeled GHI in watts per square metre
The file begins in 2009, giving over a decade of modeled reference values that can be aligned to any sub-period of the measured record. Sample rows from modelated/LQ.csv:
date,GHI
2009-01-01 09:00:00,0.0
2009-01-01 10:00:00,1.9555556
2009-01-01 11:00:00,118.24
2009-01-01 12:00:00,296.35556

Atmospheric Reanalysis Reference — modelated/LQ_regs_60.csv

A complementary file modelated/LQ_regs_60.csv provides hourly atmospheric reanalysis variables that are used as regression features in later modeling stages:
ColumnDescription
datetimeISO-8601 timestamp (hourly)
tmNear-surface temperature (normalised)
uwEastward (zonal) wind component
dwNorthward (meridional) wind component
tcwvTotal column water vapour (kg m⁻²)
d2m2-metre dewpoint temperature (K)
Sample rows from modelated/LQ_regs_60.csv:
datetime,tm,uw,dw,tcwv,d2m
2009-01-01 00:00:00,0.07969801,-0.0008236779,-0.0005172433,11.591854,280.4798
2009-01-01 01:00:00,0.07900763,-0.00051642524,-0.00033304427,11.803848,280.09528
These variables come from atmospheric reanalysis products and capture moisture, temperature and wind conditions that modulate how much solar radiation reaches the surface, even under nominally clear skies.

Dataset at a Glance

PropertyValue
StationLQ (La Quiaca area, Argentina)
Altitude3,500 m a.s.l.
Measurement period2020 – 2023
Temporal resolution1 minute
Total raw rows (all years)~1,764,898
Total years4
Modeled reference start2009
Sensor typePyranometer (real field measurements)

Continue Exploring

Loading Data

Step-by-step guide to reading, renaming, parsing and reindexing the multi-year CSV files with pandas.

Measurement Stations

Reference for all 16 GHI stations in the South American network, including geographic coordinates.

Build docs developers (and LLMs) love