Remote sensing is the acquisition of information about an object or phenomenon without making physical contact with the object. In this context it refers to the use of NASA’s orbiting satellite OCO-2 to observe characteristics of the Earth’s atmosphere, specifically the concentration of Carbon Dioxide (CO2).

The OCO-2 satellite measures the composition of the atmosphere by analysing sunlight reflected off the surface of the Earth. It primarily uses three different acquisition modes to collect data: In Nadir mode the satellite points straight down towards the Earth. In Glint mode the satellite points toward the spot where the sun is reflected at an angle from the surface of the Earth. In Target mode the satellite locks on to a specific surface location for up to 9 minutes – these target locations contain a ground-based CO2 measurement instrument that is used to calibrate the readings from the satellite.

The enormous quantity of data produced by OCO-2 is practically useless in its raw format. Before we can use it to understand our atmosphere and predict the climate, the data must go through several stages of processing.

Packet data is the raw data stream as it is sent from the orbiting satellite to a ground station. In the Level 0 data product, duplicates and transmission errors have been removed. Level 1A and Level 1B involve transformations of the data e.g. calibrations and re-sampling. The Level 2 product is derived from Level 1 data, arranged and formatted in a more accessible way for public release, however it is still split into different files for each acquisition mode and contains a high level of noise and/or error.

The Level 2 Lite data product combines the acquisition modes and includes bias correction and warning levels indicating the quality of each data point. In my Level 2 Lite visualisations below, I have excluded data points above warning level 10.

In this table of OCO-2 data visualisations from September 2014 to January 2016 you can see how the different acquisition modes have been combined to create the Lite data product. The unit is CO2 Parts Per Million in total column of atmosphere, or XCO2.

Level 2 Nadir Level 2 Glint Level 2 Lite

This data is important because it allows us to locate the major sources and sinks of atmospheric CO2, and to see how the composition of the Earth’s atmosphere is changing over time. For example we see many more readings above 400 PPM when comparing January 2015 to Jan 2016:

OCO-2 Level 2 Lite: January 2015 vs. January 2016
OCO-2 Level 2 Lite: January 2015 vs. January 2016, Atmospheric CO2 Parts Per Million