SHEFS - SDM pipeline

author: Vivienne Groner

date: 09.03.2022

source: my_path

DATA PRE-PROCESSING, SDM RUN ON CLUSTER, POST-PROCESSING

1. prepare GBIF occurence records (download, clean, rarify, create background and pseudoabsences, extract environmental data)

    01_SA_prep_occurrence_2021.R (loop over list of species in folders, my_path/04_occurrence_records/GBIF_raw)
   

2. create sh and R jobs for cluster

    02_job_maker.R
   
    SA_dummy.R
   
    SA_dummy.sh
   
    03_make_batch_files.R
   

3. run SDM on cluster (present (1979-2013) and RCP45, RCP85 for 5 GCMs)

   the following data need to be on cluster:

• species specific R and sh file
• batch files
• SA_functions_CBER.R
• my_path/02_environmental_data/CHELSA/
• my_path/02_environmental_data/CHELSA_future/
• my_path/04_occurrence_records/GBIF_ready/environmental/
• my_path/04_occurrence_records/GBIF_ready/points/

4. copy tar.zip files from myriad to RDS

    login myriad -> login xxxxxxx@transfer02
   

5. untar folders

    04_untar.R
   
    cmd < cmd_call.txt
   

6. reorder SDM output in new structure (-> I copied core folders to my_path/11_SDM_output/data/)

LAND USE AND INTENSITY

1. resample land cover data to 1 km resolution (reclassifed to PREDICTS and high level land cover classes in ArcMap)

    05_SA_NLC_2020_resample.R
   

2. create land use and intensity suitability map

   calculate effect sizes for each group (arthropods, birds, amphibia, reptiles, mammals, gastropods)

    06_PREDICTS_subset.R 
   

   create map for 2020 and 2080 ssp245 and ssp585

    07_create_LCC_scenarios_CropNatUrbWat.R
   

DATA ANALYSIS

	SA_functions_CBER_analysis.R

read SDM output and create dataframe with presence locations

	08_SDM_df_climate_current.R

	08_SDM_df_climate_future.R

and calculate LUI effect:

	08_SDM_df_climate_lui_current.R

	08_SDM_df_climate_lui_future.R

SENSITIVITY TESTS

distribution of AUCs for all species

	code: 09_sensitivity_analysis_AUC.R

sensitivity test of buffer size and density for pseudoabsences

	code: 10_sensitivity_analysis_pseudoabsence.R

output: my_path/04_occurrence_records/buffer_test/

sensitivity test LUI factors

	code: 11_sensitivity_analysis_LUI_factors.R

	output: my_path/02_environmental_data/LUI_sensitivity/