I am a meteorologist and climatologist. I began my scientific life as a student working with the famous Russian climatologist Mikhail I. Budyko and continued to work with him in the Main Geophysical Observatory and State Hydrological Institute (St. Petersburg, Russia) for many years. My first research was on the energy balance of the atmosphere and the earth-atmosphere system. Then, I participated in the Russian program on satellite observation of the Earth's radiation balance. Later I began work on the global climate change problem and was among the leaders of Russian research on greenhouse global warming. In 1975, I organized the Research Laboratory on Contemporary Climate Change at the State Hydrological Institute, St. Petersburg, Russia. My research interests at that time were monitoring of global and regional climate change, study of climate sensitivity, and the effect of global climate change on water resources. The main results of my work during that period were published in my monograph Climate Sensitivity, in which new methods of empirical study of climate sensitivity were developed.

In 1991 I accepted the invitation of Syukuro Manabe and spent the next two years working as a visiting scientist in his group at GFDL/Princeton University. My main finding during that period was that since contemporary climate models are almost as complicated as nature, the same statistical methods should be applied for both modeling results and observed data if we are going to compare them. This approach convinced me that climatic models are much more realistic than their authors think they are.

In 1993 I moved to the University of Maryland to work with Alan Robock in the Department of Meteorology. Although he moved to Rutgers University in 1998, we still work together on projects related to soil moisture and climate change detection. I was interested in change in land surface hydrology related to climate change on different scales. I have also worked on remote sensing of soil moisture using microwave observations from satellites.

As a member of the Graduate School at the State Hydrological Institute, St. Petersburg, Russia I supervised 6 Ph.D. students from 1975 to 1991: 4 in meteorology, 1 in hydrology and 1 in oceanography. Two of them, Pavel Groisman (NCDC) and John Antonov (NODC), successfully continue their scientific careers in the US. Since 1994, I have been an Adjunct Member of the Graduate School of the University of Maryland. At Maryland, I have been a research advisor, together with Alan Robock, of 3 Ph.D. students. The first of them, Adam Schlosser (at COLA now) received his Ph.D. in 1995, and the second, Jared Entin (at GSFC/NASA now), received his Ph.D. in 1998. A third student, Shuang Qiu, received her M.S. in 1998 and began work for NESDIS/NOAA.

My research was always between numerical models and observations. I always worked with global or large-scale data sets of conventional and satellite-retrieved data to study the ocean-land-atmosphere climatic system. The most significant subject of my current scientific interest is the cryosphere (sea ice and snow cover) and its interaction with atmospheric and hydrospheric processes. More details about my research directions are:

• Terrestrial energy balance and climate sensitivity. For many years I worked on the energy balance of the earth-atmosphere system and on empirical study of climate sensitivity. Although I am not currently actively working on these topics, they both are still in the sphere of my research interests. I have a particular interest in the energy budget of polar regions and better understanding of the effects of sea ice and snow cover variations on climate sensitivity.
• Global and regional climate change. In 1976, I first detected a century scale 0.5K/100 yr warming trend in surface air temperature averaged over the Northern Hemisphere and recognized that this trend was related to greenhouse global warming. M. I. Budyko and I published a paper then, entitled "Global warming," thus launching these two words into the modern world. Later, I used better data and a statistically optimum method for global averaging of climatic records and developed, together with Pavel Groisman and Kira Lugina, an original method of monitoring mean global temperature changes. Our data have been used in many scientific reports on global climate change problem, including the prestigious IPCC reports, and I was among the lead authors of the IPCC (1990, 1992) reports. I currently work together with Alan Robock and GFDL scientists to estimate the probability of observed climatic trends to appear as the result of natural climate variability.
• Soil moisture. The GFDL scientists predicted that global warming may be responsible for summer desiccation of continents, but it was unclear to what extent this model prediction was realistic. I made a paleoclimatic reconstruction of the soil moisture regime for a few warm periods of the past, which showed that changes of soil moisture in response to global warming are much more complicated. In this work I came to understand that the research community needed in situ soil moisture data to validate the models, but no data were available. I started to work with such data in Russia and began to use them for model validation and scale analysis. Later Alan Robock and I organized a special project to gather, make available, and study soil moisture observations, and found that many other countries, including the US, China, Mongolia, and India have a history of long term soil moisture observations. A significant part of these data is now available from the electronic Global Soil Moisture Data Bank that we created. They are in use for model validation in the GSWP/ISLSCP, PILPS 2(d), AMIP, and LDAS projects.
• Scales of atmospheric and land surface processes. There is a significant difference in the temporal and spatial scales of atmospheric and land surface processes. This is the reason that meteorologists, hydrologists, and remote sensing scientists sometimes find it difficult to understand each other?s approaches to study of the same system. The global network of meteorological stations was designed to be representative of atmospheric forcing and to avoid the influence of land surface variability in such things as topography, vegetation, and soils. An increase of spatial resolution of a satellite sensor permits more and more information to be obtained about the landscape-related variability of an observed variable. For example, for the midlatitude soil moisture field, the landscape related variability is equal to up to 50% of the soil moisture field variance, and for this part of the variance, the scale of spatial correlation is about 10 m and scale of temporal autocorrelation is a few days. The other part of soil moisture variability is related to atmospheric forcing, and has a spatial scale of a few hundred km and temporal scale of a few months. Our instruments and observational networks are always responsible for some filtration of the original signal. Together with Alan Robock, I currently work to study statistical properties and scales of atmospheric and land surface processes, their interaction, and the optimization of observational networks and satellite remote sensing systems.
• Remote sensing of soil moisture. The theory of passive microwave remote sensing of soil moisture is well developed but it contains too many unknown physical parameters of soil, land surface, and vegetation. Therefore, microwave satellite indices must be calibrated using direct measurements of soil wetness. Data analysis convinced me that SMMR and SSM/I observations can be used to retrieve the component of soil moisture variability that is related to atmospheric forcing. The results were just published in JGR, co-authored with Alan Robock, Bhaskar Choudhuri, Eni Njoku, and students. I believe that work on calibration of SMMR and SSM/I passive microwave observations to create a satellite-based soil wetness data set should be continued.
• Sea Ice. My first attempt to detect a global warming signal in the sea ice extent record was done almost two decades ago. At that time it was impossible to distinguish the very weak signal from the noise. A few years ago, I began to analyze the GFDL climate model-predicted change in Northern Hemisphere sea ice related to greenhouse global warming. I organized an informal scientific team that includes specialists on sea ice (John Walsh, Victor Zakharov), snow cover (Alan Robock, David Robinson) and climate modeling (Ronald Stouffer). I received updated satellite retrieved sea ice data sets from NCEP/NOAA (Chet Ropelewski and Don Garrett), the GSFC/NASA group (Claire Parkinson and Don Cavalieri), and from Norwegian scientists (Einar Bjorgo and Ola Johannessen). The results have been published in Science. These results are almost shocking. The GFDL climate model being forced by CO₂ and sulfate aerosols has revealed that very significant changes in the area of very thick sea ice should have happened during the second half of this century. It is very possible that up to half of sea ice thicker than 3 m has been lost and this loss has not yet been noticed. The observed trend in NH sea ice extent is in satisfactory agreement with the model-predicted trend, but it looks as if the observed trend exceeds the model predicted one. I used a very long control run of the same climate model to assess the natural variability of sea ice extent and found that observed changes have a very low probability to be the result of natural climate variability. I am currently work on sea ice problem in collaboration with NASA scientists Don Cavalieri and Claire Parkinson.
• Trend analysis.   Seasonal and diurnal cycles can be found in averages and climatic trends of climatic records.  A few years ago I began to work on a new statistical technique that can be applied to records with such cycles.  The approach and details of the technique have been described in several recent publications.  My coathors are Alan Robock, Don Cavalieri, Claire Parkinson, Norman Grody, Alan Basist.  This technique produces unprecedented opportunities to analyze and visualize climatic change.  There is no restriction that the climatic process be stationary, and records can contain seasonal and diurnal cycles in the moments of the statistical distribution (means, variances, and higher moments) of the observed variables and their trends.  The technique can be applied to historical observations of meteorological stations with many changes in their times of observation, and can also be applied to climate model output.  It It has been applied to analysis of trends of  50-year records of surface air temperature at a several US stations, showing that decreases in diurnal temperature range occur only in summer and fall.  It has been applied also to analyze of trends in satellite-observed sea ice, showing that the Northern Hemisphere trend is much different from that in the Southern Hemisphere.  The most important was analysis of the trend in satellite-observed tropospheric temperature, showing that the upward trend of tropospheric temperature for the past 25 years is the same as the trend in surface temperatures.  Previous analyses showing different trends, and claims that that result invalidated global warming from surface temperatures and predictions of global warming from climate models, are no longer justified.  The last result has been published in Science.
  

• Collaboration with other scientists and research centers. I have many fruitful interactions with Alan Robock at Rutgers University, Ronald Stouffer at NOAA's Geophysical Fluid Dynamics Laboratory, John Walsh at the University of Illinois, Norman Grody at NESDIS/NOAA, Claire Parkinson and Don Cavalieri at GSFC/NASA, and several other people actively working in the remote sensing and climate change modeling areas.