Partitioning of memory and real-time connections between variables in Himalayan ecohydrological process networks

Abstract

The Himalayan ecosystem is a global biodiversity hotspot and a vital component of the global water cycle. However, the studies characterizing the ecohydrological processes of the Himalayas are still limited. Looking at a system as a network, having nonlinear couplings, can give us better insights into its dynamics. Here, using an information-theoretic approach on the variables, Precipitation (P), Temperature (T), Enhanced Vegetation Index (EVI), Latent Heat Flux (LH), Sensible Heat Flux (SH), Wind Speed (WS), Incoming Shortwave Radiation (SWL), and Relative Humidity (Q), we represent the ecohydrological processes of the Himalayas in the form of networks for three seasons: summer (MAM), monsoon (JJASO), and winters (NDJF). The networks have two types of links between variables: real-time and memory-driven. We show that the couplings between ecohydrological variables in the Western Himalaya are more memory dominant that the Eastern Himalaya. Precipitation interacts with vegetation in the Himalayas using both real-time associations as well as memory-based connections. The dominance of memory varies spatially and temporally. The Temperature, on the other hand, influences vegetation in near real-time, and it also has memory-based links in Central Himalaya and at the higher elevations of the Eastern Himalaya. We find that the real-time interactions (zero lagged connections) among ecohydrological variables are high during the monsoon as opposed to winters, which are dominated by memory-based associations. These findings provide the foundation for further analysis of the trajectory of Himalayan ecohydrological systems under natural and human-induced climate stresses.