Global Weather Station Investment Analysis by Decisive Data
10-31-2017 11:37 AM
10-31-2017 11:37 AM
With over 7,200 Weather Stations around the globe, it is imperative that investors and public officials gauge these station’s consistency and geographic presence. Doing so will lead to better preparation of future network health through resource allocation, closer monitoring, or other strategies.
The Global Weather Station Investment Analysis Dashboard is designed to help decision makers identify which temperature measurement stations around the world are consistently outputting data, as well as to identify stations which are isolated enough for their reporting data to be considered more valuable than others.
Stations send in their minimum, maximum, and average temperature each month. With this data, found at https://www.ncdc.noaa.gov/ghcnm/, we can extract, transform, load, analyze, and eventually visualize the value of each individual weather station around the world. Stations with missing or invalid measurements have a lower consistency score, and stations that are clustered together have a lower isolation score.
Isolation Score Calculation: 1 / Number of Stations within 200km
Consistency Score Calculation: Valid measurements sent / All Measurement Periods
Recommended Station Action falls into four different categories:
Stations that both Isolated and High-Consistency need to be sustained, as they are the most valuable type of station.
Stations that are isolated but have low-consistency should be supported. Supporting them to increase performance will elevate them to the most valuable classification.
Stations that are not isolated but are still classified as high-consistency should be acknowledged. While there are no actions these stations can take to become ‘isolated’, they are encouraged to remain high-consistency in the event that stations nearby are deactivated.
Stations that are non-isolated and low-consistency should be deactivated and support, resources, and funding should go to stations that fall under the ‘support’ classification.
- Stations can be inactive. As default, only ‘active’ stations are given recommended actions. However, stations can also be classified as ‘inactive for 10 years’ and ‘inactive for 20+ years’ for historical analysis.