Interannual variability of regional evapotranspiration under precipitation extremes: A case study of the Youngsan River basin in Korea

- We assessed basin-scale evapotranspiration (ET) in annual precipitation (P) extremes.
- The regional ET decreased in the severely large-P year and increased in small-P year.
- Limiting factor in ET was the available energy during summer under monsoon climate.
- Leaf area index was the key factor to determine spatial distribution of ET responses.
- The regional ET was invariable in terms of water balance, regardless of P extremes.

SummaryUnderstanding basin-scale evapotranspiration (ET) is an important issue for the management of regional water resources, especially with the recent trend of intensified precipitation (P). This study assessed the spatial and temporal variations of regional ET in response to P extremes, for various types of land-cover across the Youngsan River basin in Korea.The spatial distribution of monthly P and ET from 2001 to 2009 were estimated using rainfall records from 40 weather stations located across the basin and a satellite-derived, process-based ET model Breathing Earth System Simulator (BESS) (Ryu et al., 2011), respectively. The study periods were focused on the recent years with abnormally large, small and normal P, which were identified from anomalies in the z-sores of long-term (1973-2011) rainfall records. The variation of regional ET was assessed in terms of: (1) the controlling factors, using the statistics of related meteorological and geographical data, (2) a water-energy balance, using Budyko's framework, and (3) the water balance of four selected watersheds in the region, using the partitioning of annual P into ET and riverflow discharge (Q).The total annual ET of this region decreased in abnormally large-P year and increased in small-P year, because the ET in July to August (which accounts for more than 36% of annual ET) was limited by the available energy rather than available water due to the summer monsoon. In terms of land cover types, forests showed larger interannual variability in ET than paddy fields or cropland, with the differences in ET between large and small-P years being 108 and 82 mm yr−1, respectively. The sensitivity of annual ET to P extremes was significantly related to the leaf area index (LAI), rather than soil properties, topography, or specific land-cover type (p < 0.05, generalized linear model). However, the interannual variations of ET were not large (15-18%) compared to those of annual P (51-62%) and Q (108-232%) during 2002-2009. Thus, vegetation played a consistent role in releasing water back to the regional atmosphere through ET, regardless of P extremes.