#98 updated references

Author: alizma

paper/paper.bib

@@ -28,6 +28,13 @@ @software{Liew:2022
 version={1.0.0}
 }
 
+@software{hawc2:2007, 
+author={DTU Wind}, 
+year={2007}, 
+title={HaWC2: An Aeroelastic Code}, 
+url={https://www.hawc2.dk/}, 
+}
+
 @article{mann_spatial:1994,
 	author = {Mann, Jakob},
 	doi = {10.1017/S0022112094001886},
@@ -43,7 +50,7 @@ @article{mann_spatial:1994
 
 @article{mann_wind:1998,
 	author = {Mann, Jakob},
-	doi = {https://doi.org/10.1016/S0266-8920(97)00036-2},
+	doi = {10.1016/S0266-8920(97)00036-2},
 	issn = {0266-8920},
 	journal = {Probabilistic Engineering Mechanics},
 	number = {4},

paper/paper.md

@@ -32,7 +32,7 @@ Synthetic turbulence models (STMs) are used in wind engineering to generate real
 
 `DRDMannTurb` aims to create an easy-to-use framework to (1) fit one-point spectra from data using the DRD model introduced in [@Keith:2021] and (2) efficiently generate synthetic turbulence velocity fields to be used by scientists and engineers in downstream tasks. Existing methodologies for generating synthetic turbulence frequently incur a large computational overhead and lack the DRD model's flexibility to represent the diverse spectral properties of real-world observations, cf. [@Liew:2022]. `DRDMannTurb` addresses these two issues by introducing (1) a module for fitting neural network-based DRD models to observed one-point spectra data as well as (2) a module for efficiently generating synthetic turbulence boxes. Rather than generating turbulence for the entire box at once, which is a highly memory-intensive practice used in other software, `DRDMannTurb` uses a state-of-the-art domain decomposition approach to generate smaller sub-boxes sequentially.
 
-`DRDMannTurb` is completely written in Python, leveraging computationally powerful backend packages (`numpy`, `PyTorch`). Our implementation allows for easy GPU-portability using `cuda`. This is an additional advantage compared to previously developed software packages that have implemented the Mann model but do not provide the source code (e.g., HAWC2). Finally, `DRDMannTurb` is designed to be more general-purpose, allowing it to be applied to a broader range of scenarios and making it more accessible with clear documentation for a variety of tasks that researchers in this area are often interested in. 
+`DRDMannTurb` is completely written in Python, leveraging computationally powerful backend packages (`numpy`, `PyTorch`). Our implementation allows for easy GPU-portability using `cuda`. This is an additional advantage compared to previously developed software packages that have implemented the Mann model but do not provide the source code (e.g., HAWC2 [@hawc2:2007]). Finally, `DRDMannTurb` is designed to be more general-purpose, allowing it to be applied to a broader range of scenarios and making it more accessible with clear documentation for a variety of tasks that researchers in this area are often interested in. 
 
 # Results