Effect of COVID19 Aerosol Reduction on Rainfall in the Western United States

2.1. Study Area and Sample Description

The study was carried out in the Western United States, a region with complex terrain and seasonal differences in rainfall. Ninety-five meteorological stations were selected from federal and state networks. Stations were included only if more than 90% of daily records were available from January 2018 to December 2021. The sample covers both coastal valleys and mountain ranges such as the Sierra Nevada and the Rocky Mountains. This setup allows comparison between regions with warm-rain processes and those with mixed-phase systems.

2.2. Experimental and Control Design

Observations from March to December 2020 were used as the experimental period. This time corresponds to the lockdown when emissions from traffic and industry were reduced. The control period was defined as the same months in 2018, 2019, and 2021, when no major emission changes occurred. Using the same months reduced seasonal effects, and including multiple years reduced random year-to-year variation. The main assumption is that the largest difference between the two periods comes from aerosol changes.

2.3. Measurement Methods and Quality Control

Daily rainfall data were taken from the National Centers for Environmental Information (NCEI) and state climate networks. Aerosol optical depth was obtained from the MODIS Terra and Aqua satellites. Before analysis, all datasets were checked. Stations with gaps longer than 10 consecutive days were removed. Outliers were detected with the interquartile range method and compared with nearby stations. Satellite data followed the standard NASA cloud-screening procedure. These steps ensured that both ground and satellite records were consistent.

2.4. Data Processing and Model Equations

Data were converted to monthly values to reduce short-term variation. A linear regression model was applied to test the relation between rainfall and aerosols [14]:

$${\mathrm{R}}_{\mathrm{t}}=\mathsf{\alpha }+{\mathsf{\beta }}_1\mathrm{AO}{\mathrm{D}}_{\mathrm{t}}+{\mathsf{\beta }}_2{\mathrm{T}}_{\mathrm{t}}+{\mathsf{ϵ}}_{\mathrm{t}}$$

where ${\mathrm{R}}_{\mathrm{t}}$ is rainfall in month $\mathrm{t}$, $\mathrm{AO}{\mathrm{D}}_{\mathrm{t}}$ is aerosol optical depth, ${\mathrm{T}}_{\mathrm{t}}$ is mean temperature, and ${\mathsf{ϵ}}_{\mathrm{t}}$ is the error term.

A rainfall anomaly index was also calculated [15]:

$$\mathrm{RAI}={\displaystyle \frac{{\mathrm{R}}_{\mathrm{obs}}-{\mathrm{R}}_{\mathrm{mean}}}{{\mathrm{R}}_{\mathrm{mean}}}}\text{×}100\%$$

where ${\mathrm{R}}_{\mathrm{obs}}$ is rainfall during the study period and ${\mathrm{R}}_{\mathrm{mean}}$ is the average rainfall of the control years.

2.5. Statistical Analysis

Statistical work was performed in R (version 4.1). Correlation coefficients were used to measure the link between aerosols and rainfall. Significance was tested at the 95% confidence level. Analyses were repeated separately for mountain and valley stations. A sensitivity test was done by removing one year at a time from the control period to check the stability of results. This reduced the effect of single-year extremes and increased the reliability of the findings.

3.1. Lockdown Changes in Aerosols and Related Pollutants

During March–May 2020, records from California showed a decline in PM2.5 and a reduction in ozone compared with the 2015–2019 average. The change was not uniform. Coastal valleys and the Central Valley showed the largest decreases in PM2.5, while some cities had smaller or mixed ozone changes. These results indicate that the lockdown reduced key pollutants, although AOD did not fall everywhere [16]. Figure 1 shows the spatial differences in California during the lockdown.

3.2. Regional Precipitation Signals During the Lockdown Window

Rainfall during the lockdown months showed small changes with clear regional differences. Valleys where warm-rain processes are common recorded slight increases. Mountain regions with orographic and mixed-phase clouds showed little change. These results agree with microphysical theory: fewer condensation nuclei may increase droplet size and enhance coalescence in warm clouds, but mixed-phase systems respond in more complex ways [16]. Figure 2 summarizes the physical pathways that explain these outcomes.

3.3. Site-Scale Links Between AOD and Precipitation

Regression results using AOD, temperature, and rainfall showed weak relationships at most sites. Where significant, the slopes were negative in mountain stations during winter and positive in some valley sites during summer. These different signs are expected because AOD measures column particles, while cloud-active nuclei depend on type, height, and humidity [17,18]. Local weather and storm features also affect rainfall, which limits the ability to explain rainfall changes by AOD alone. The schematic in Figure 2 supports this view by showing how aerosols affect warm and mixed-phase clouds in different ways.

3.4. Context, Limits and Implications

The lockdown created a short but clear reduction in emissions, yet rainfall changes were small and depended on terrain and cloud regime. Two limits are important. First, short study periods reduce statistical strength and make it hard to separate weak signals from natural variation. Second, AOD does not fully describe cloud-active particles, which differ by type and condition. Using maps of pollutant changes (Figure 1) and a schematic of cloud processes (Figure 2), the results suggest that cleaner air may slightly enhance warm-rain formation in valleys. However, circulation and terrain remain the main drivers of rainfall in the western U.S.