TL;DR
This paper uses a high-resolution mesoscale model (RAMS) to simulate diurnal atmospheric circulation patterns over the complex terrain of the Valencia coastal region, Spain. By comparing model outputs with extensive ground-based and airborne measurements, the study demonstrates the model's ability to accurately capture key circulation regimes, including nocturnal drainage, sea breezes, and their interactions with topography, providing confidence in using such models for understanding pollutant transport.
What Problem Does This Paper Address?
Many of the world's population lives in coastal areas, often characterized by complex terrain. In these regions, atmospheric gases and aerosols emitted at the surface are transported by mesoscale circulations, which are not resolved by global transport models. These mesoscale circulations, driven by diurnal solar heating and influenced by local topography, play a crucial role in the transport and distribution of pollutants. Understanding these processes is vital for accurate air quality forecasting and for linking emission sources to observed atmospheric concentrations. The Valencia basin, with its coastal plain backed by mountains, presents a challenging environment for studying these phenomena. This paper aims to characterize and simulate the diurnal mesoscale circulation regimes in this region and evaluate the performance of a mesoscale model in capturing these complex dynamics.
Methods
The study employs the Regional Atmospheric Modeling System (RAMS) with a high-resolution, nested domain configuration (down to 1.5 km grid spacing) to simulate atmospheric conditions over the Valencia basin. The model incorporates advanced schemes for turbulence parameterization (Mellor and Yamada level 2.5), radiation (Chen and Cotton), sub-grid scale convection (Kuo-modified), and a LEAF-2 land surface scheme for calculating heat and moisture fluxes. A key methodological aspect was the initialization of soil moisture and temperature using a 2-month spin-up period, which was found to improve model performance compared to using global analysis data directly. The model's simulations were then rigorously evaluated against a comprehensive dataset of ground-based and airborne meteorological measurements collected during a field campaign.
Datasets
The study utilizes data from a 2-week intensive measurement campaign conducted in the Valencia basin during July 2001. This campaign involved:
- Ground-based measurements: Meteorological data (temperature, humidity, wind speed and direction) from seven surface weather stations (PO, CA, SA, RI, AL, VI, XA). Eddy covariance flux measurements of CO2, H2O, and energy were also collected at two towers (El Saler and a rice field site).
- Aircraft measurements: Vertical profiles and horizontal transects of meteorological parameters (sensible and latent heat fluxes, temperature, humidity, wind) collected using a Sky Arrow ERA aircraft over rice fields.
- Meteorological analysis data: NCEP Aviation (AVN) global model analysis data was used for initial and boundary conditions for the RAMS model. NOAA-AVHRR raw satellite data was used for sea surface temperature.
Key Insights and Ideas
- Diurnal Circulation Regimes are Dominant: The Valencia basin exhibits distinct diurnal circulation patterns: nocturnal katabatic drainage winds channeled by valleys, a daytime sea breeze that merges with upslope winds, and an evening transition phase where larger-scale pressure gradients can influence the flow.
- Terrain-Driven Transport is Crucial: The complex topography significantly shapes the wind patterns, leading to localized convergence and divergence zones that influence vertical transport and the development of pollution layers.
- Sea Breeze Interaction with Mountains: The sea breeze doesn't just penetrate inland; it interacts with the mountain slopes, leading to upslope winds and subsequent injection of air into the atmosphere, followed by a compensating subsidence over the sea.
- Model Captures Key Dynamics: Despite the inherent complexity and uncertainties (e.g., soil moisture initialization, turbulence parameterization), the RAMS model, particularly after a spin-up period, successfully simulates the timing and intensity of these diurnal circulation regimes, including the inland progression of the sea breeze.
- Soil Moisture Initialization Matters: A 2-month spin-up for soil moisture and temperature significantly improved model performance compared to direct initialization from global analysis data, though it still introduced some regional biases (wet inland, dry coastal).
Key Results
- Temperature: The model accurately captured the diurnal temperature range at stations, with a slight under-prediction at the coastal station SA. The warming trend between July 1st and 2nd was well reproduced. The overall temperature RMSE was 1.49°C.
- Specific Humidity: The model generally simulated higher specific humidity than observed, indicating a wet bias, especially inland during the day. The RMSE for specific humidity was 3.5 g kg⁻¹.
- Winds: The model showed good agreement with observed wind patterns, capturing the transition from nocturnal drainage to daytime sea breeze. The RMSE for wind speed was 1.54 ms⁻¹, and the Index of Agreement (IOA) was 0.79. However, discrepancies were noted in wind direction during nocturnal transitions and at inland sites. The Vector Wind Difference (VWD) RMSE was generally between 1 and 3 ms⁻¹, with higher values inland.
- Circulation Patterns: Visualizations of horizontal streamlines and vertical cross-sections clearly depicted the nocturnal drainage, the development of the sea breeze, and the associated vertical transport and subsidence.
Limitations and Future Work
- Soil Moisture Biases: The model exhibited a dry bias near the coast and a wet bias inland, potentially impacting the simulation of fluxes and the early stages of diurnal cycles. This is attributed to the simple soil moisture initialization and potentially limitations in the LEAF-2 scheme.
- Turbulence Parameterization: The paper suggests that the RAMS model's stable boundary layer parameterization might over-diffuse the flow during nocturnal regimes, leading to overly homogeneous wind fields and potentially more mixing than observed. This could be a limitation for accurately simulating pollutant dispersion.
- Spatial Resolution: While high-resolution, the model's resolution might still be insufficient to fully resolve fine-scale humidity profiles and the extreme variability of wind observed in complex terrain.
- Future Work: The authors suggest that more experimental data, particularly over complex terrain and at higher altitudes, would be beneficial for further validating and improving mesoscale models. Investigating the impact of turbulence parameterizations on tracer transport is also highlighted as an important area for future research. The companion paper explores the application of this model to CO2 transport.
Summary updated