Add doi to article.

Author: erlenbs

README.md

@@ -2,7 +2,7 @@
 
 This repository contains a complete set of scripts and notebooks applied to estimate the *Expected Annual Damage* (EAD) to roads asociated with river floods (in Portugal). The framework may be adapted to other regions and other types of hazards, if represented by some hazard intensity parameters associated with a damage function. To estimate the uncertainty associated with the EAD, the damage is modeled as a spatial random field. The spatial correlation of this random field is decisive to the uncertainty of the EAD. To this end, Gaussian Processes are used to embed the uncertainty associated with the flood damage function with a spatial correlation.
 
-The methodology is described in the article *Uncertainty in flood risk assessment of linear structures: Why correlation matters*.
+The methodology is described in the article [*Uncertainty in flood risk assessment of linear structures: Why correlation matters*](https://doi.org/10.1016/j.jhydrol.2023.130442).
 
 ## Data
 All the data used for the analysis is freely available. Where to find it, and how to download it is described below.
@@ -126,4 +126,4 @@ The support is gratefully acknowledged.
 ## Further development.
 The current framework does not estimate uncertainty with respect to the flood intensity parameters. The framework should be adapted to also consider uncertainty associated with flood maps.
 
-Some straight OSM road segments have large distances between their spatial coordinates. A refinment step to obtain an upper bound on distance beetween segment coordinates may be implemented as part of the `assign_raster_to_osm_elements.py` script.
+Some straight OSM road segments have large distances between their spatial coordinates. A refinment step to obtain an upper bound on distance beetween segment coordinates may be implemented as part of the `assign_raster_to_osm_elements.py` script.