A Hybrid Approach for Precipitation Simulation Using the Multi-site Neyman-Scott Rainfall Pulse Model and the Analog Method
Abstract
The Multi-site Neyman-Scott Rainfall Pulse model (MNSRP) was previously implemented in many studies over different climate zones and is known for its capability to accurately capture the statistical and the geostatistical properties of the precipitation process up to the third moment, which is a key factor in extreme precipitation analysis. However, due to the purely stochastic nature of the model, larger spatial precipitation patterns cannot be captured, especially when the model is implemented over large regions. The Analog Method (AM) uses observed precipitation EOFs to accurately simulate precipitation spatially. The quality of the AM simulations suffers in arid and semi-arid regions where the statistical properties of the precipitation process are harder to capture. The Hybrid Approach (HA) uses the AM simulations to locally rank the MNSRP simulations for each month. Case studies in different climatic regions in the US show that the new approach compares favorably to each individual model.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2011
- Bibcode:
- 2011AGUFM.H21D1140K
- Keywords:
-
- 1854 HYDROLOGY / Precipitation;
- 1910 INFORMATICS / Data assimilation;
- integration and fusion;
- 1984 INFORMATICS / Statistical methods: Descriptive;
- 4311 NATURAL HAZARDS / Analogue modeling