A method to calculate a locally relevant water year for ecohydrological studies using eddy covariance data: Locally Relevant Water Year

Increased rainfall variability due to climate change significantly impacts carbon and water cycling in ecosystems, but these impacts may be masked when using arbitrary annual reporting periods such as the calendar year, which may not have any relevance to natural annual ecosystem processes. A variety of alternative annual integration periods have been described for specific purposes or locations, but are of limited general applicability. Here, we present an eddy covariance data-driven empirical method to determine a locally relevant annual time period. The method selects a start date for a locally relevant water year (LRWY) that maximizes correlation between annual precipitation (AP), and annual evapotranspiration (AET) and annual gross primary production. The method was tested using data from 2004 to 2013 for 2 Ameriflux sites (woodland and grassland) in Central Texas. The timeframe included periods of unusually high rainfall and periods of extreme drought. The highest correlation between AP, and AET and annual gross primary production was obtained with an LRWY beginning in mid-September. Use of the LRWY better captured the impact of soil water recharge in the autumn on photosynthesis the following spring than did calendar years. Use of the LRWY also identified more annual periods in which AET exceeded AP, which more accurately reflected the impact of drought on ecosystems processes than did analysis based on calendar years.