Prophets of global warming are winning new respect from skeptical scientists. Subtle changes in weather patterns that conform to predictions of improved computer-climate simulations strongly suggest that man-made climate change is upon us.
As one former skeptic, Thomas Karl, puts it: ''I think there's a likelihood that the changes we see are not just due to natural variability. There is a human component.''
These changes include such effects as a greater share of precipitation coming in winter and more precipitation coming in extreme events, more severe warm-season droughts, an increase in above-normal temperatures, and a decrease in day-to-day variability in temperature in parts of the Northern Hemisphere.
The draft of a new assessment by the Intergovernmental Panel on Climate Change (IPCC) likewise concludes that the observed 0.5 degree C rise in average global temperature over the past 135 years ''is unlikely to be entirely due to natural causes.''
Dr. Karl - a senior scientist with the National Climatic Data Center in Asheville, N.C. - says that right now, we're at a point where a majority of scientists in the field believe there may be a global-warming signal. But many are not yet willing to say that it is an obvious signal. He expects that to change to a near-consensus in favor of global warming within five years if present weather trends continue.
Michael MacCracken, with the United States Global Change Research Program office in Washington, makes a similar point. He says the evidence is ''quite compelling'' that heat-trapping so-called greenhouse gases and the cooling effect of man-made aerosols are affecting climate. Yet it is so far not enough to convict human activity of climate change ''beyond all reasonable doubt.''
The most important of these aerosols are sulfates emitted by such sources as coal- and oil-fired power plants. Like particles from volcanoes, they block sunshine and cool the lower atmosphere. Computer models had trouble simulating past climate trends in this century - let alone predict the future - when they took account only of the buildup of such greenhouse gases as carbon dioxide and methane. But when aerosols are included, the simulations become much more realistic.
For example, this was the key to temperature changes that puzzled James Hansen, Makiko Sato, and Reto Ruedy at the Goddard Institute for Space Studies in New York. The average daily maximum temperature over land rose by 0.28 degrees C between 1951 and 1990. Average minimum temperature rose 0.84 degrees - three times as much. Yet computer models with only carbon dioxide forcing predicted the same warming day and night. To match both the long-term 0.5 degree C global warming and the observed changes in daily maximum and minimum temperatures, the researchers had to include the cooling influences of aerosol pollution and increases in mid-level cloudiness.
One possible indicator of climate warming is the decrease in Arctic sea ice reported last July in Nature magazine by Ola Johannessen, Martin Miles, and Einar Bjorgo at the Nansen Experimental and Remote Sensing Center in Bergen, Norway. They report that data from satellites show sea ice shrinkage of about 2.8 percent from 1978 to 1987 and about 4.5 percent from 1987 to 1994. They say such polar ice-cover reductions ''could provide an early indication of greenhouse warming.''
Meanwhile, Mike Hulme at the University of East Anglia in Norwich, England has found rainfall shifts in rain-gauge records across the Northern Hemisphere. In this century, rainfall has declined south of latitude 23.4 degrees, changed little between latitudes 23.4 and 30 degrees, and increased north of 50 degrees. That's just what computer simulations predict.
Then there are the latest findings reported by Dr. Karl and his colleagues. They explained in Nature last month that day-to-day temperature variability has decreased in the Northern Hemisphere. Moreover, they report that ''at least within the US - the proportion of total precipitation contributed by extreme, one-day events [more than 2 inches rain equivalent per day] has increased significantly.'' That amounts to ''an average of about one additional extreme precipitation event every two years,'' the scientists say. Here again, they note that these findings are ''in agreement with model projections of a warmer world.''
There is less certainty as to what global warming will bring. According to a study released last month by the Global Change Research Program, it's ''virtually certain'' the stratosphere will cool. It's ''very probable'' that global mean surface temperature will rise by about 0.5 to 2 degrees C between 1990 and 2050. It's ''very probable'' that global mean precipitation will increase, with changes in distribution that are harder to foresee. And it's ''very probable'' that Arctic land areas will warm in winter. Global sea-level rise is also ''very probable,'' although its extent is uncertain.
It is ''probable'' that Northern Hemisphere summer droughts will increase as will high-latitude precipitation. It also is ''probable'' that Antarctic and North Atlantic ocean regions will warm more slowly than the global average temperature.
For his part, Karl says, ''My conviction is that it is indeed going to get warmer.'' He adds that, over the next five years, he expects that people generally ''will begin to notice a climate change.''