predict_per_tiles: Derive daily space-time predictions at standard soil depths

In kanj241/agrosoil: Surface and Rootzone Soil Moisture Prediction

Derive daily space-time predictions at standard soil depths

Description

Derives daily gridded distribution of soil moisture at standard soil depth intervals for a list of spatial tiles.

Usage

predict_per_tiles(
  i,
  model_,
  in_path,
  out_path,
  targ_var,
  method,
  days_,
  month_,
  month.lst_,
  pred.year_,
  rds_file,
  DDepth,
  stdps = c(5, 15, 30, 60, 100),
  weather_data = FALSE,
  weather_source = "ERA5",
  era5_depth = FALSE
)

Arguments

i	Vector of spatial tile IDs
model_	Saved model (random forest or xgboost models) to be used for spatial predictions on the larger grid or stacked tile grids (.rds file)
in_path	Directory holding the stacked spatial tiles
out_path	Directory holding the stacked spatial tiles. Just to keep predictions with their respective stacked covariates.
targ_var	Target variable to predict
method	Machine learning algorithm to be used for the prediction. Values are "ranger" and "xgboost". Default value is "ranger" random forest algorithm.
days_	Numeric. Days in the respective month of the targetted soil moisture prediction year. Default values are all days in the respective month of the year
month_	Character. Single month of the targetted soil moisture prediction year.Default value is the months of the in situ soil moisture meansurement. Values are abbreviated month name. E.g. "Jan", "Mar", "Oct".
month.lst_	Vector or a list. All months of the targetted soil moisture prediction year.
pred.year_	Numeric. Single soil moisture prediction year
rds_file	Saved stacked-spatial tile of covariates (i.e. prediction grid)
DDepth	TRUE/FALSE. Whether to predict for top soils (FALSE) or for specific standard soil depths (TRUE).
stdps	Vector. Numeric values of standard soil depths at which predictions are required. Default values are c(0,5,15,30,60,100,200)

Examples

library(rgdal)
library(raster)
library(lubridate)
library(plyr)
library(caret)

out <- predict_per_tiles(i, model_, in_path, out_path, targ_var, method, days_, month_, month.lst_,pred.year_, rds_file, DDepth, stdps)

kanj241/agrosoil documentation built on March 25, 2022, 12:22 a.m.