Peer review of Bakker et al 2016
The Bakker et al 2016 paper focuses on a Python-based tool called FloPy, which is meant specifically for creating predictive models of groundwater flow. The uses of FloPy are presented as an alternative for commonly used GUIs (Graphic User Interface) to predict the spread of ground water, taking into account many different environmental factors that may impact it. The paper takes time to explain all mechanisms within the code, step-by-step, and thoroughly explains the logic of each line contained within the code. It uses specific examples from the script, as well as several graphs and charts to highlight the potential advantage of using the seemingly old fashioned scripting method. I myself have zero training in hydrology, and my peer review of this paper can only be done from the perspective of a novice Python user. My personal interest lies in the fact that the scripting mechanisms used in creating a predictive model for groundwater flow in FloPy could be used for predictive modeling across various other disciplines.
In the introduction, Bakker et al 2016 describe in short the mechanisms behind a typical GUI. The process is presented clearly, and understandable even without much prior knowledge about hydrology or predictive modeling. This helps quite a bit in being able to really understand the advantages of FloPy over the currently more popular GUIs. The paper also explains that the Python language was chosen specifically, because it has a very powerful syntax, and many tools at the programmers disposal (thus earning its popularity in scientific circles.)
The next portion of the study discusses a specific example of a FloPy use. The authors do a great job explaining the function of the tool step-by-step, which allows for a thorough understanding of its functions. As with every script, there are a set of attributes to be defined at the very beginning, including but not limited to; geographical boundaries, water level depth, and periodical water fluctuation. Specific instructions are provided for how varying metrics ought to be filled out in order for the model to run appropriately. The next part of the paper focuses on more advanced applications of FloPy, and describes how individual scripters can go about adding unique metrics that would need to be considered in specialized settings that would apply to more complicated study areas. Several specific studies that used the FloPy tactic are cited and highlighted, which in its own way serves as the testing of the papers initial hypothesis.
In conclusion, the advantages of using FloPy over current GUI methods are several. First, the time of processing the model tends to be twice as fast. Second, a script automatically provides a record of each specific run, which can itself be peer reviewed, modified and re-tested. Third, the capabilities of Python allow for a more specialized predictive model than currently available with the more popular GUIs. The Bakker et al 2016 paper presented a hypothesis clearly, demonstrated that it had been tested on several occasions, and drew logical conclusions based on the results of said tests.
Link:
https://hostsited2l.uwf.edu/content/enforced/1024185-51499GIS5103201805/SupplementalReadings/PotentialsForPeerReview/Bakker_et_al-2016-Groundwater.pdf?_&d2lSessionVal=OhmXFvyMsrT2qJz3PCvegYrVo&ou=1024185
