diff --git a/README.md b/README.md
index 59bf0e8..5bd129a 100644
--- a/README.md
+++ b/README.md
@@ -19,9 +19,9 @@ Please cite it when using this repository for your own studies.
 ## Workflow
 1. Derive daily maxima from GESLA3 data and write them to CSV files using the script [daily_max_GESLA3_to_csv.py](https://github.com/Timh37/TimingAFs/blob/main/GPD_analysis/daily_max_GESLA3_to_csv.py).
 2. Load the CSV files and apply the method of Solari et al. (2017) to the daily maxima using [gpdfit_solaris_thres_gesla3.R](https://github.com/Timh37/TimingAFs/blob/main/GPD_analysis/gpdfit_solaris_thres_gesla3.R). This script automatically selects the extreme threshold and outputs central-estimate general Pareto distribution parameters and their uncertainty to NetCDF files. A script to compute distribution parameters for a constant threshold of 98.8% is also included ([get_gpd_parameters988.R](https://github.com/Timh37/TimingAFs/blob/main/GPD_analysis/get_gpd_parameters988.R)). The script [ats_output_to_dataset.py](https://github.com/Timh37/TimingAFs/blob/main/GPD_analysis/ats_output_to_dataset.py) is used to merge the NetCDF files containing the automatic threshold selection output for each location into a single NetCDF file for all locations, keeping only the data needed for the timing projections.
-3. To project the timing of AFs, [project_AF_timing_ar6wfs.py](https://github.com/Timh37/TimingAFs/blob/main/project_timing/project_AF_timing_ar6wfs.py)  takes the GPD parameter estimates and fetches the FLOPROS estimates nearest to the GESLA3 locations. Using the GPD fits, the return curves and required SLR is computed. The timing of AFs is then projected by interpolating the timing of projected SLR onto required SLR, and all is stored in a single NetCDF file for all locations.
+3. To project the timing of AFs, [project_AF_timing_ar6wfs.py](https://github.com/Timh37/TimingAFs/blob/main/project_timing/project_AF_timing_ar6wfs.py)  takes the GPD parameter estimates and fetches the FLOPROS estimates nearest to the GESLA3 locations. Using the GPD fits, the return curves and required SLR are computed. The timing of AFs is then projected by interpolating the timing of SLR projected by the IPCC AR6 onto the required SLR derived from the return curves, and the resutls are stored in a NetCDF file containing the projections for all locations.
 
-4. (optional) Scripts to reproduce Figures 4 and 5 of the main manuscript can be found [here](https://github.com/Timh37/TimingAFs/blob/main/plotting). The timing projections used for these plots are available at [Zenodo repository]().
+4. (optional) Scripts to reproduce Figures 4 and 5 of the main manuscript can be found [here](https://github.com/Timh37/TimingAFs/blob/main/plotting). The timing projections used for these plots are available at [this Zenodo repository]().
 
 <p align="right">(<a href="#top">back to top</a>)</p>