Rainfall Forecasting Skill over the Tropical Andes

Abstract:

The interaction between tropical synoptic-scale disturbances and the Andes mountains play a critical role in NW South America weather and climate. The main goal of this work is to evaluate rainfall forecast from a one year-long, cloud resolving weather forecasting simulations using a very dense network of raingauges, and remote sensing products from the Tropical Rainfall Measurement Mission (TRMM) and Global Precipitation Measurement (GPM) mission. The forecasts are based on the Weather and Research Forecasting (WRF) model at 4 km grid size and driven with the Global Forecasting Systems (0.5) forecast data for 72 hours and initialized daily for the 00 and 12 UTC cycles. The model day-ahead forecasting performance, and systematic biases are evaluated as a function of phase and amplitude of the westerly and easterly flow synotic disturbances. The model’s ability to produce episodic and month-long rainfall topoclima patterns shed light on the levels and sources of synoptic predictability in the region, with easterly wind anomalies related to the Tropical Easterly Waves troughs dominating predictablity of synoptic wet/dry spells in the region. Additionally, we further show that a 4 km, subdaily, and unbiased monitoring rainfall system based on a hybrid between day-ahead rainfall forecasts, raingauges, near real-time/coarser-scale rainfall satellite estimates outperforms other rainfall analysis products.