A fork of {cropsim} (Hijmans et al. 2009) designed to make using the EPIRICE model (Savary et al. 2012) for rice diseases easier to use and to implement the EPIWHEAT model in the same R package. This package provides easy to use functions to fetch weather data from NASA POWER, via the {nasapower} package (Sparks 2018) and predict disease intensity of five rice diseases using a generic Susceptible-Exposed-Infectious-Removed (SEIR) model (Zadoks 1971) function, seir().
This package fully implements the EPIRICE model first published in Savary et al. (2012), which introduces the model and uses it to model global epidemics of rice diseases illustrating the risk of bacterial blight, brown spot, leaf blast, sheath blight and rice tungro disease. Additional models included for rice are adapted from Kim et al. (2015) for rice sheath blight and leaf blast. The EPIWHEAT model first published in Savary et al. (2015) that was used to model potential epidemics of two wheat diseases, leaf rust and Septoria tritici blotch, is also included in this package.
The modern core of seir() in {epicrop} uses a different numerical integration approach than the legacy SEIR() in {cropsim}. While both implement the canonical SEIR formula (diseased = latent + infectious + removed), the modern version uses cumulative delay windows instead of forward Euler integration. This results in different dynamics and typically lower AUDPC values. Additionally, the modern seir() is 1-indexed (standard R convention) while the legacy SEIR() is 0-indexed (as in the original EPIRICE publication by Savary et al. 2012).
Scripts are available in inst that benchmark and validate this version against the original EPIRICE. As the original EPIWHEAT code is not available, no such comparisons are possible. I have recreated this as closely as possible to the original, while fixing minor bugs and implementing more efficient routines. In benchmarking this iteration is up to 57% faster in seasonal disease calculations than the original EPIRICE. This varies by OS and hardware, but the gains should be notable across large datasets, e.g., several years of global data as in Savary et al. 2012.
Users may also provide their own parameters to the generic seir() function to simulate other crop diseases as long as the weather data provided meet the model’s requirements.
Quick start
{epicrop} is not yet on CRAN. You can install it this way.
install.packages(
"epicrop",
repos = c("https://adamhsparks.r-universe.dev", "https://cloud.r-project.org")
)Get weather data
First you need to provide weather data for the model; {epicrop} provides the get_wth() function to do this. Using it you can fetch weather data for any place in the world from 1983 to near present by providing the and latitude and dates or length of rice growing season as shown below.
library(epicrop)
# Fetch weather for year 2000 wet season for a 120 day rice variety at the IRRI
# Zeigler Experiment Station
wth <- get_wth(
lonlat = c(121.25562, 14.6774),
dates = "2000-07-01",
duration = 120
)
wth
#> Key: <YYYYMMDD>
#> YYYYMMDD DOY TEMP TMIN TMAX RHUM RAIN LAT LON
#> <IDat> <int> <num> <num> <num> <num> <num> <num> <num>
#> 1: 2000-07-01 183 25.29 23.86 27.78 92.20 23.12 14.6774 121.2556
#> 2: 2000-07-02 184 26.13 23.54 29.90 86.01 17.34 14.6774 121.2556
#> 3: 2000-07-03 185 25.50 24.28 27.23 94.16 29.08 14.6774 121.2556
#> 4: 2000-07-04 186 25.81 24.50 27.56 92.42 13.01 14.6774 121.2556
#> 5: 2000-07-05 187 25.97 25.13 27.40 92.34 32.20 14.6774 121.2556
#> ---
#> 117: 2000-10-25 299 25.82 23.44 29.54 89.76 12.04 14.6774 121.2556
#> 118: 2000-10-26 300 25.44 24.14 26.99 94.93 13.03 14.6774 121.2556
#> 119: 2000-10-27 301 25.74 24.54 27.69 91.43 11.54 14.6774 121.2556
#> 120: 2000-10-28 302 25.44 24.72 26.62 91.90 74.20 14.6774 121.2556
#> 121: 2000-10-29 303 24.97 24.15 26.34 94.15 29.11 14.6774 121.2556Modelling bacterial blight disease intensity
Once you have the weather data, run the model for any of the five rice diseases by providing the emergence or crop establishment date for transplanted rice.
bb_sim <- bacterial_blight(wth, emergence = "2000-07-01")
bb_sim
#> simday dates sites latent infectious removed senesced rateinf rlex rtransfer rremoved rgrowth
#> <int> <IDat> <num> <num> <num> <num> <num> <num> <num> <num> <num> <num>
#> 1: 1 2000-07-01 100.0000 0 0.0000 0.000 1.000000 0 0 0 0.00000 9.68750
#> 2: 2 2000-07-02 108.6875 0 0.0000 0.000 2.086875 0 0 0 0.00000 10.49959
#> 3: 3 2000-07-03 118.1002 0 0.0000 0.000 3.267877 0 0 0 0.00000 11.37416
#> 4: 4 2000-07-04 128.2934 0 0.0000 0.000 4.550811 0 0 0 0.00000 12.31499
#> 5: 5 2000-07-05 139.3254 0 0.0000 0.000 5.944065 0 0 0 0.00000 13.32593
#> ---
#> 116: 116 2000-10-24 0.0000 0 396.0528 1404.565 2665.037553 0 0 0 43.67864 0.00000
#> 117: 117 2000-10-25 0.0000 0 356.3610 1444.257 2712.120592 0 0 0 39.69181 0.00000
#> 118: 118 2000-10-26 0.0000 0 309.2779 1491.340 2755.012575 0 0 0 47.08304 0.00000
#> 119: 119 2000-10-27 0.0000 0 266.3860 1534.232 2791.854071 0 0 0 42.89198 0.00000
#> 120: 120 2000-10-28 0.0000 0 229.5445 1571.073 2824.358958 0 0 0 36.84150 0.00000
#> rsenesced diseased intensity lat lon
#> <num> <num> <num> <num> <num>
#> 1: 1.000000 0.000 0 14.6774 121.2556
#> 2: 1.086875 0.000 0 14.6774 121.2556
#> 3: 1.181002 0.000 0 14.6774 121.2556
#> 4: 1.282934 0.000 0 14.6774 121.2556
#> 5: 1.393254 0.000 0 14.6774 121.2556
#> ---
#> 116: 39.691810 1800.618 1 14.6774 121.2556
#> 117: 47.083039 1800.618 1 14.6774 121.2556
#> 118: 42.891982 1800.618 1 14.6774 121.2556
#> 119: 36.841497 1800.618 1 14.6774 121.2556
#> 120: 32.504887 1800.618 1 14.6774 121.2556Lastly, you can visualise the result of the model run.
library(ggplot2)
ggplot(
data = bb_sim,
aes(
x = dates,
y = intensity
)
) +
labs(
y = "Intensity",
x = "Date"
) +
geom_line() +
geom_point() +
theme_classic()
Bacterial blight disease progress over time. Results for wet season year 2000 at IRRI Zeigler Experiment Station shown. Weather data used to run the model were obtained from the NASA Langley Research Center POWER Project funded through the NASA Earth Science Directorate Applied Science Program.
Meta
Please report any issues or bugs.
License: GPL-3
To cite {epicrop}, please use the output from
citation(package = "epicrop").
Code Coverage
#> epicrop Coverage: 95.99%
#> R/get_wth.R: 77.22%
#> R/interpolation.R: 80.34%
#> R/validation.R: 84.62%
#> R/format_wth.R: 91.18%
#> R/leaf_wetness.R: 97.54%
#> R/audpc.R: 97.83%
#> R/run_epicrop_model.R: 99.26%
#> R/seir.R: 99.32%
#> R/build_epicrop_emergence.R: 100.00%
#> R/disease_models.R: 100.00%
#> R/epi_get_params.R: 100.00%
#> R/fetch_epicrop_weather_list.R: 100.00%Code of Conduct
Please note that the epicrop project is released with a Contributor Code of Conduct. By contributing to this project, you agree to abide by its terms.
Other Implementations in R
The EPIRICE model was originally written in R as a part of the {cropsim} package (Hijmans et al. 2009).
The EPIRICE model is also available from CRAN in the {ZeBook} package (Brun et al. 2018) to accompany, “Working with dynamic crop models: methods, tools and examples for agriculture and environment” (Wallach et al. 2018).
References
François Brun, David Makowski, Daniel Wallach, and James W Jones. (2018). ZeBook: Working with Dynamic Models for Agriculture and Environment. DOI: 10.32614/CRAN.package.ZeBook R package version 1.1, https://CRAN.R-project.org/package=ZeBook.
Kwang-Hyung Kim, Jaepil Cho, Yong Hwan Lee, and Woo-Seop Lee. (2015). Predicting potential epidemics of rice leaf blast and sheath blight in South Korea. Agricultural and Forest Meteorology, 203: 191-207. DOI: 10.1016/j.agrformet.2015.01.011
Robert J Hijmans, Serge Savary, Rene Pangga and Jorrel Aunario. cropsim. (2009). Simulation modeling of crops and their diseases. R package version 0.2-6.
Serge Savary, Andrew Nelson, Laetitia Willocquet, Ireneo Pangga and Jorrel Aunario. (2012). Modeling and mapping potential epidemics of rice diseases globally. Crop Protection, Volume 34, Pages 6-17, ISSN 0261-2194 DOI: 10.1016/j.cropro.2011.11.009.
Serge Savary, Stacia Stetkiewicz, François Brun, and Laetitia Willocquet. Modelling and Mapping Potential Epidemics of Wheat Diseases-Examples on Leaf Rust and Septoria Tritici Blotch Using EPIWHEAT. (2015). European Journal of Plant Pathology 142, no. 4:771–90. DOI: 10.1007/s10658-015-0650-7.
Adam Sparks. (2018). nasapower: A NASA POWER Global Meteorology, Surface Solar Energy and Climatology Data Client for R. Journal of Open Source Software, 3(30), 1035, DOI: 10.21105/joss.01035.
Jan C Zadoks. (1971). Systems Analysis and the Dynamics of Epidemics. Phytopathology 61:600. DOI: 10.1094/Phyto-61-600.
Wallach, Daniel, David Makowski, James W Jones, and François Brun. (2018) Working with dynamic crop models: methods, tools and examples for agriculture and environment. Academic Press. DOI: 10.1016/C2016-0-01552-8.