What's the difference between 'climate data' and 'climate data platforms'?
There's only limited ways to monitor the weather and climate. Raw data might come from satellites, computer models, rainfall radars or weather stations.
Climate & weather products or 'weather data'
The data-sources above get made into a range of different products. For example, satellite rainfall products include ENACTS, CHIRPS, TAMSAT and ARC2. Each product will have different strengths and weaknesses in different places and for different uses. For example, CHIRPS (a satellite product that merges in rain-gauges) is better at capturing high rainfall events than CHIRP (which doesn't), but it takes longer to become available. A weather product will have the following attributes:
- A spatial resolution (E.g. a 10km grid, or a point source at this location)
- A temporal resolution (e.g. daily, monthly)
- Some numbers which correspond to rainfall/temperature etc for each spatial location and time-stamp.
- Some way of recording missing values
Climate data statistics
There are lots of ways you can use these weather products. For example, you might sum a daily satellite product to look at the total seasonal rainfall over a point, or over a region, or use the data to calculate the number of dry days. Some products above will be better than others at making these different statistics.
Data Platforms and tools
These are the tools that you use to view the climate statistics above. One tool may show multiple statistics, from multiple products, or one product may crop up across a lot of tools. A specific tool might let you analyse a climate product in a certain way, for example it might allow you to calculate your own statistics, or easily zoom in on a point. The IRI Data Library, or Princeton Drought Monitoring portal are examples of platforms.
So you cannot say if a platform is 'good' or 'bad'. You can just comment on whether you like the way that it analyses the data.
For example, the Princeton portal cuts off Northern Somalia in some of its plots. The underlying TRMM data product does exist for northern Somalia, but the portal is simply unable to show it.
An example of the same data product (NOAA RFE2), viewed as default on different data platforms. RFE2 is a product made out of raw data from Meteosat (Infra-red) satellite data, AMSU (Passive microwave) satellite data and limited rain-gauges. It is a daily product available from 2002 until the present day and is created on a 10km grid.
Three different data platforms incorporate the RFE2 product. Each platform shows its own chosen statistics as default (GIEWS = 10-daily sums over Somalia, IRI data library = daily data over your region of choice (with the ability to create more statistics) and the NOAA website just the most recent 10-daily sum.
So when we assess climate data, we can assess:
- The skill of RFE2 to accurately capture rainfall over Somalia
- The different statistics that you might want from it (e.g. it is more skilful for regional rainfall totals compared to daily rainfall at a specific pixel)
- The properties of the data portal. E.g. I like the colour scale of the GIEWS viewer, but I wish it would allow me to zoom in to an individual region.
Note to Dustin - this would make a cool infographic