The European heatwave of 2018 exhibited record-breaking temperatures and extreme dry conditions across the whole continent. Our study considers novel and very recent solar-induced fluorescence (SIF) data acquired from the Orbiting Carbon Observatory (OCO-2) satellite. We explore SIF variation during spring and summer months across different vegetation types (agriculture, broad-leaf forest, coniferous forest, and mixed forest) during the European heat and drought in 2018 and compare it to non-drought conditions.

SIF is proposed to be a direct proxy for gross primary productivity (GPP) and thus can be used to draw inferences about changes in photosynthetic activity in vegetation due to extreme events. About one-third of Europe experienced a consecutive spring and summer drought and about one-sixth a severe spring and summer drought in 2018. A definite pattern change in SIF variation is observed within the spring season for all vegetation types, with lower SIF during the start of spring followed by increased in fluorescence from mid-April, even though the SIF difference from the mean was not significant. Summer, however, showed a significant decrease in SIF with the values constantly lower than the mean value. Our results show that particularly agricultural areas were severely affected by the drought and heat of 2018. Further, the variation in the spring-summer SIF was captured across different Köppen climate zones in Europe. The intense heat wave in Central Europe showed about 31 % decrease in SIF value during July and August as compared to the mean over three years.

NASA’s Orbiting Carbon Observatory (OCO-2) satellite measurement is one of the most used datasets in the literature of greenhouse gases (GHG). This satellite measures the total column averaged CO₂ concentration (also referred to as XCO₂) as contiguous parallelogram footprints, roughly one kilometer in diameter. One of the major hindrance associated with this dataset is the missing XCO₂ footprints in one particular time instance which eventually propagates in its time-series measurements.

There are few other global datasets related to GHG measurements which mainly provide high-resolution anthropogenic CO₂ emission information worldwide, such as ODIAC, EDGAR, FFDAS. This information is highly related to the atmospheric CO₂ concentration variation and if modeled as the auxiliary information for the prediction of missing CO₂ concentration footprints, may enhance the accuracy in prediction. In this work, we are aiming to combine these emission inventory information along with OCO-2’s XCO₂ data for its interpolation process. A Semantic Kriging based interpolation method is considered to carry out the prediction. This method is capable to incorporate related auxiliary information (both numerical and semantic) of the terrain into the prediction process for the enhancement of the prediction accuracy. A motivating example scenario and corresponding source code can be found here.