King-Fai Li, PhD
Assistant Professor of Environmental Sciences/Statistics,
University of California, Riverside
EMAIL: (at)




Appointments Education Academic Awards Research Interests Research Descriptions

My research interests broadly cover various fields in atmospheric science, from Earth to exoplanets, from tropical dynamics, stratospheric chemistry to radiative transfer, on processes of time scales ranged from diurnal to decadal, or to even billion years. These works require knowledge in statistics including data retrieval and signal processing, and modeling such as sensitivity to input parameters. My long-term interest is to utilize ground-based and satellite observations and to evaluate our capability in predicting climate change during the 21st century with state-of-the-art climate models.

Solar cycles and Stratospheric Chemistry

Whether the solar cycle has important influence on the current climate remains as a controversial issue. Nonetheless, it is sometimes used to test the climate sensitivity of general circulation models. Since most of the solar cycle variations occur in the shortwave region (~120-300 nm), any solar-cycle effects on climate are expected to be enhanced in the middle atmosphere via solar heating due to stratospheric ozone absorption. The different heating of the stratosphere as a function of latitude due to these absorption processes may modify the tropospheric circulation, leading to changes in the hydrological cycle. Thus thorough knowledge of the chemical and thermodynamic changes in the stratosphere is a key for understanding how solar cycle may relate decadal changes in our climate. One of my research interests is to characterize the stratospheric changes related to the solar cycle in temperature, ozone, and other chemical species from both models and ground-based/satellite observations. Recently, I resolved a 30-year puzzle about the 11-year solar response of tropical ozone, pointing out that a part of the ozone decadal variability in the long-term satellite record was contaminated by satellite orbital drifts and that conventional photochemistry explains the ozone solar response satisfactorily after removing the effects of the satellite orbital drifts. At high latitudes, I showed that the solar cycle may act to increase the polar ozone by ~20% by weakening the polar vortex, thereby increasing the inflow of ozone-rich air from low latitudes to the polar region, confirming the teleconnection mechanism of the solar cycle modulation at low latitides. These work have been summarized in

Intraseasonal Variability of Atmospheric Chemical Tracers

The Madden-Julian oscillation (MJO) is the most dominant form of intraseasonal variability in the tropical atmosphere. The dynamical aspects of MJO have long been known and are subjects of active research. However, the impact of the MJO on atmospheric compositions has been realized only recently. One of my research projects examined how MJO modulates the distribution of carbon dioxide (CO2) in the tropics. CO2 is the most important anthropogenic greenhouse gas in the present-day climate. To date, almost all of the discussions on atmospheric CO2 have focused on the variability with time scales from annual to centennial time scales, either because of the limitations of data or the relevant time scales that relates climate change. Recently, mid-tropospheric CO2 retrieved by the NASA Atmospheric Infrared Sounder (AIRS) provides us the first global daily CO2 data, allowing the study of subseasonal variability. In 2010, I found a subseasonal variability in the 7-year data of AIRS CO2 associated with the Madden-Julian oscillation (MJO). The signal is about ~1 ppmv or ~0.3% of the mean CO2 in the atmosphere, which is of the critical scale for identifying both oceanic and land sources of carbon flux. I also showed evidence that the modulation is likely to be driven by lower tropospheric vertical motions. These short-term processes can organize, transport and mix CO2 and provide a robustness test for coupled carbon-climate models. I have also been examining the MJO-modulations in other tracers such as water vapor, ozone, and carbon monoxide. The MJO can also modulate other tracers. The detailed structures of these MJO impacts are useful for chemical transport modelers and will help improve the forecasts of atmospheric compositions. Papers related to these research activities include

Possible astrobiological signatures on exoplanets

Currently the only method for detection of extraterrestrial life is by remote sensing. A recent discovery of water vapor, methane and carbon dioxide on an exoplanet suggests that habitable exoplanets might be very common in the universe. The remaining question is whether these exoplanets might have life. According to Drake's equation, the probability of detecting an extraterrestrial civilization is directly proportional to the duration over which they broadcast (e.g. in electromagnetic waves). Such duration is obviously bounded by the life span of their biosphere. In Gaia Hypothesis, the Earth's biosphere, through the regulation of the global surface temperature (in geological time scales) by the carbonate-silicate cycle around the evolving Sun, can last only for ~1 Ga more from present. In 2009, my colleagues and I suggested that the global surface temperature may also be regulated by the atmospheric pressure. This would potentially extend the life span of the biosphere to ~2 Ga, doubling the chance of humans being detected by extraterrestrial intelligence. The change in the atmospheric pressure may lead to signatures in infrared spectrum, which may be observable with current or near-future technologies. This research has been discussed in

Advanced statistical methods for climate studies

Secular climate changes are generally small and are difficult to detect. To examine these secular changes, much larger climate variations such as the seasonal cycle and the El Nino have to be removed first. Linear statistical methods such as the correlation analysis are usually employed for the removal but the processes to be removed are required to be stationary and linear, which is generally not true for real data. There have been growing interests in developing adaptive methods that decompose a non-linear time series based only on the nature of the time series per se, in the way that the decomposition does not require any assumptions or prior information. Examples of such adaptive methods include the Empirical Mode Decomposition (EMD), and the Decomposition Matching Pursuit (DMP). In 2012, I worked with two summer undergraduates and used the above two methods to examine the climate signals imprinted in atmospheric temperature and chlorofluorocarbons (CFCs). We successfully extracted a less-known near-annual oscillation (period ~1.5 yr) in the atmospheric temperature and a change in the seasonal cycle of the CFCs after the global CFC regulation in 1995. These studies demonstrate the usefulness of these newly developed statistical methods for climate studies. Papers related to these studies are



















Selected Conference Presentations

Last Modified: PST