diff --git a/docs/paper.bib b/docs/paper.bib
index 088546e0..def1dc61 100644
--- a/docs/paper.bib
+++ b/docs/paper.bib
@@ -1,4 +1,4 @@
-@article{Brady2021, 
+@article{Brady:2021, 
     doi = {10.21105/joss.02781}, 
     url = {https://doi.org/10.21105/joss.02781}, 
     year = {2021}, 
@@ -10,7 +10,7 @@ @article{Brady2021
     journal = {Journal of Open Source Software} 
     } 
 
-@article{https://doi.org/10.1002/met.1732,
+@article{Griffiths:2019,
 author = {Griffiths, Deryn and Foley, Michael and Ioannou, Ioanna and Leeuwenburg, Tennessee},
 title = {Flip-Flop Index: Quantifying revision stability for fixed-event forecasts},
 journal = {Meteorological Applications},
@@ -50,7 +50,7 @@ @article{Ehm2016
  year = {2016}
 }
 
-@article{Taggart2022,
+@article{Taggart:2022,
 author = {Taggart, Robert and Loveday, Nicholas and Griffiths, Deryn},
 title = {A scoring framework for tiered warnings and multicategorical forecasts based on fixed risk measures},
 journal = {Quarterly Journal of the Royal Meteorological Society},
diff --git a/docs/paper.md b/docs/paper.md
index 19f49f22..f458fcdb 100644
--- a/docs/paper.md
+++ b/docs/paper.md
@@ -43,7 +43,7 @@ The research purpose of this software is (a) to mathematically verify and valida
 
 ## Related Works
 
-`climpred` [@Brady2021:2021] provides some related functionality and provides many of the same scores. `climpred` does not contain some of the novel functions contained within `scores`, and at the same time makes some design choices specifically associated with climate modelling which do not generalise as effectively to broader use cases as may be needed in some circumstances. Releasing `scores` separately allows the differing design philosophies to be considered by the community.
+`climpred` [@Brady:2021] provides some related functionality and provides many of the same scores. `climpred` does not contain some of the novel functions contained within `scores`, and at the same time makes some design choices specifically associated with climate modelling which do not generalise as effectively to broader use cases as may be needed in some circumstances. Releasing `scores` separately allows the differing design philosophies to be considered by the community.
 
 `xskillscore` [@xskillscore] provides many of the same functions as `scores`. `xskillscore` does not contain some of the novel functions contained within `scores` and does not contain the Jupyter Notebook tutorials which provide users with clear guidance on the use of the verification metrics.