Sensitivity of rainfall characteristics to thermodynamic and dynamic vertical structure over Costa Rica
Abstract
Understanding how the vertical aspects of thermodynamic and dynamic structure relate to rainfall in the tropics is critical for interpreting observed variability and refining representations of rainfall in numerical weather prediction and climate models. Here, multivariate K-means clustering is performed on vertical profiles of moisture, temperature, and zonal and meridional winds from four decades of WMO radiosoundings from San Jose, Costa Rica as a basis for identifying a small set of representative vertical structures that are used for analyzing regional rainfall characteristics. By systematically varying the number of clusters through a plausible range (k = 2 to 6), it is shown how aspects of regional seasonality and large-scale controls on the convective environment, such as the dominant wet vs. dry season, the semi-annual strengthening of the Caribbean Low-Level Jet (CLLJ) and the July-August occurrence of the Midsummer Drought (MSD), impact the profiles of moisture, temperature, and winds. The cluster assignments are used to composite simple convection-relevant indicators such as the lifting condensation level, convective available potential energy, and convective inhibition to highlight systematic variation of these diagnostics under distinct vertical thermodynamic and dynamic structures. Cluster assignments are further used to composite large-scale meteorological fields from the ERA5 reanalysis to demonstrate the physical consistency of the clustered vertical structures with the larger-scale environment. Particular emphasis is devoted to identifying the signatures of easterly waves (EWs) in the clustering, as EWs are known to be important high-frequency drivers of rainfall over Costa Rica, though questions remain about precisely how EWs affect vertical thermodynamic and dynamic structure over land and the interplay of that vertical structure with rainfall. Overall, the cluster-based analysis enables detailed characterization of the large to local scale relationships inherent to tropical rainfall and facilitates interpretation of rainfall events in the context of the interaction of regional features such as CLLJ and MSD and higher frequency drivers like EWs over land.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2020
- Bibcode:
- 2020AGUFMH141.0023W
- Keywords:
-
- 3305 Climate change and variability;
- ATMOSPHERIC PROCESSES;
- 1817 Extreme events;
- HYDROLOGY;
- 1854 Precipitation;
- HYDROLOGY;
- 4318 Statistical analysis;
- NATURAL HAZARDS