rpicsa: R Package for PICSA and Climatic Analysis for Crops

Overview

rpicsa is an R package designed to assist in the analysis of climate data for crop production using the Plant Impact on Climate and Societal Adaptation (PICSA) framework. It provides a range of functions to help researchers, agronomists, and farmers make informed decisions about planting and managing crops based on climatic data.

The package offers a suite of tools to perform various climatic analyses, including:

• Start of Rains: Determine the onset of the rainy season, an important factor in crop planning.
• End of Rains: Identify when the rains conclude, helping in crop harvesting and preparation for the dry season.
• End of Season: Identify when the rainy season concludes, helping in crop harvesting and preparation for the dry season.
• Length of Season: Calculate the duration of the growing season, a useful parameter for crop selection.
• Rainfall Summaries: Generate summaries of rainfall within a rainy season to understand precipitation variability.
• Temperature Summaries: Calculate temperature statistics to assess the impact on crop development.
• Crop Success Probability: Estimate the likelihood of successful crop production based on historical climate data.
• Season Start Probability: Predict the probability of the rainy season’s onset to optimise planting schedules.

Installation

You can install the development version of rpicsa like so:

# Install the 'devtools' package if you haven't already
if (!require(devtools)) {
  install.packages("devtools")
}

# Install 'rpicsa' from GitHub
devtools::install_github("IDEMSInternational/rpicsa")

Usage

We give some small examples in using the functions in rpicsa

library(rpicsa)
library(dplyr)

# use the data in the package, but for this demonstration we will look at the data before 1950.
daily_niger <- rpicsa::daily_niger %>% dplyr::filter(year <= 1950)

To calculate the start of rains in Niger, we can use the start_rains() function. In our case, we define the start of rains as the first instance where there is more than 25mm of rainfall over three days, and there is no more than nine days without rainfall within 21 days.

We also calculate the end of rains. Here, we define the end of rains to be the first instance after day 121 (April 30th), where there is one day with no more than 10mm of rainfall.

# Calculate Start of Rains
start_rain_table <- start_rains(data = daily_niger, date_time = "date", station = "station_name",
            year = "year", doy = "doy", rain = "rain",
            total_rainfall = TRUE, over_days = 3, amount_rain = 25,
            dry_spell = TRUE, spell_interval = 21, spell_max_dry_days = 9
            )

# Calculate End of Rains
end_rain_table <- end_rains(data = daily_niger, date_time = "date", station = "station_name",
            year = "year", doy = "doy", rain = "rain",
            start_day = 121, interval_length = 1, min_rainfall = 10)

# Combine into a table Start of Rains
summary_rains_table <- dplyr::full_join(start_rain_table, end_rain_table)

We can then use this to calculate other summary statistics - for example, length of season, or rain summaries in the season

# Length of season
season_length_table <- rpicsa::seasonal_length(summary_rains_table, start_date = "start_rains", end_date = "end_rains",
                                               data = daily_niger, date_time = "date", station = "station_name",
            year = "year", doy = "doy", rain = "rain")

# We can also look at the total rainy days and rainfall amount in each season
season_rain_table <- rpicsa::seasonal_rain(summary_rains_table, start_date = "start_rains", end_date = "end_rains",
                                               data = daily_niger, date_time = "date", station = "station_name",
            year = "year", doy = "doy", rain = "rain")

We can use this to calculate probabilities to aid in successful crop production in the future. Let’s say we are just interested in Birni N'Konni. Say we want to find the best crop to plant, so try out different planting days and planting periods to see which would historically be most likely to succeed. We look at 100mm rainfall for a crop where the start of rains starts before day 100, and then before day 200, and where the length of season is 60 days or 80 days:

crops_definitions(data = daily_niger %>% dplyr::filter(station_name == "Birni N'Konni"), date_time = "date", station = "station_name", year = "year", doy = "doy", rain = "rain",
                  season_data = summary_rains_table %>% dplyr::filter(station_name == "Birni N'Konni"), start_day = "start_rains", end_day = "end_rains",
                  water_requirements = 100,
                  planting_dates = c(100, 200),
                  planting_length = c(60, 80))

We can see that the most successful crop will be one that needs a planting length of 60 days, and can be planted after day 200.

We can finally look at the probability of the rainy season to start after a specified set of dates:

probability_season_start(summary_rains_table %>% dplyr::filter(station_name == "Birni N'Konni"), station = "station_name", start_rains = "start_rains", specified_day = c(100, 200, 300))

In Birni N’Konni (from 1945-1950), the rainy season started between day 100 and 200 about 83.3% of the years. This can be useful to aid in planting schedules.

Documentation

For detailed information on each function, including arguments, usage, and examples, please refer to the package documentation. You can access it using the following command:

?rpicsa

Contributing

We welcome contributions from the community. If you have any bug reports, feature requests, or would like to contribute code to rpicsa, please visit our GitHub repository.

License

This package is open-source and distributed under the GNU Lesser General Public License v3.0 License. Feel free to use, modify, and distribute it as needed.

Contact

If you have any questions, suggestions, or feedback, please feel free to reach out to us.