Trying to figure out what is happening to the money supply and what it means these days, a Federal Reserve official complains, is "like looking into a bowl of muddy water . . . . If the outside world is confused, it should be. So are we."

Confusion notwithstanding, the Fed's policymaking group, the Federal Open Market Committee, must meet this week to set the legally required targets for money growth for 1983 that Chairman Paul A. Volcker will report to Congress on Feb. 16.

When the 19 members of the FOMC gather here Tuesday afternoon, they will confront a series of serious policy dilemmas:

First, in 1982, the normal links between money growth and the economy loosened considerably. Strong, above-target money growth failed to prevent the economy from sinking deeper in recession. Now no one can be certain whether this link--known as the velocity of money--will continue to weaken, simply reestablish itself or perhaps tighten sharply as it has on some occasions.

Second, the money supply is leaping upward, largely but not entirely as a result of a 1982 law allowing financial institutions to create two new types of deposits, the Super NOW account and the Money Market Deposit Account. In a scant six weeks, the public has shifted an incredible $213 billion into the MMDA's, while the Super NOW's have garnered another $17 billion. No one knows quite what to make of the impact of these regulatory changes on what have been the Fed's key policy targets, the measures of money known as M1 and M2.

Third, long-term interest rates are substantially higher relative to inflation than they were at the beginning of any of the other postwar recoveries. Fed officials believe they will fall even if short-term rates, which the central bank can influence much more directly, stay at or close to their present levels. But again, no one is sure. Nor are they sure what would happen if the Fed chose to push short rates down again, as it has by several percentage points since last summer.

Finally, there is the prospect of $200 billion to $300 billion federal budget deficits, if spending is not reduced and taxes raised. Such huge levels of borrowing by the government absorbs a large share of total savings. When the recovery gets rolling, and private credit demands revive, the scramble for funds could send interest rates on another upward spiral.

For the year ahead, however, the Fed's situation is not nearly as bleak as it has been for the last two or three years when its money growth targets left little room for accommodation of both real economic growth and the then-current rates of inflation. Now inflation is down to the point that the money growth targets the Fed is likely to set this week should be consistent with a recovery even if inflation were to go back up a bit--if the velocity of money behaves itself.

The Federal Reserve does want a recovery. But it wants economic growth to remain sufficiently modest--the 4 percent annual rate projected by the Reagan administration seems about right to a number of Fed officials--that inflation might continue to fall, or at least not increase significantly anytime soon.

In approaching the "specifics" of setting this year's money targets, Volcker told the Joint Economic Committee recently, "We are, and will continue to be, concerned with maintaining a monetary environment consistent both with continuing progress against inflation and with lasting expansion.

"Reconciling those goals, at a time when institutional and economic factors have called into question the reliability of past relationships between money and the economy, will be a difficult and delicate job," Volcker said. "The approach cannot be reduced to an arithmetic, or econometric, formula, nor can success be achieved by monetary policy alone."

With those and other statements, the chairman has been trying to counter any notion that the Fed is intent on achieving a recovery at all costs. Other Fed officials, equally sensitive, point out that the central bank has not lowered its discount rate, now 8.5 percent, since before Christmas because long-term rates did not fall and stay down after the last half-point reduction. Fed Gov. Nancy Teeters said last week that she favors lowering the rate again, but that she does not have the votes on the board to do it.

Some financial market participants do seem to fear that the Fed is pushing too hard to get the recovery going.

Leif Olsen, a monetarist economist who is chairman of Citibank's economic policy committee, recently told a New York bankers group that the Federal Reserve had "restored a good deal of its lost credibility in fighting inflation by pursuing a resolutely restrictive course for monetary growth in 1981 and the first half of 1982.

"But in July of 1982--and especially after the October relaxation of the money growth rates--policy has shifted onto a highly stimulative course, one that is now causing consternation in the money market," Olsen declared. "Correction for the monetary excesses of the past six months is likely to be neither smooth nor painless."

Federal Reserve officials reject Olsen's contention that money growth was too rapid in the last half of 1982. Sticking tightly to the monetary targets set early in the year would have meant an excessively tight policy because, for whatever reason, "households and firms decided to hold more money in relation to income and transactions than before," says Fed Gov. Henry C. Wallich.

Most monetarist economists maintain that the relationship between money and the gross national product, expressed in terms of current dollars, is quite stable over long periods of time. And when the velocity of money behaves strangely, as it did last year, it will settle down again quickly, they argue.

But last year's experience has unsettled some of the monetarists, including Robert E. Weintraub, an economist for the Joint Economic Committee. At a recent monetary policy conference here sponsored by the Cato Institute, Weintraub noted that the velocity of M1 fell more than 2 percent last year rather than increasing at the 3.6 percent average rate of the previous 10 years.

(M1 includes currency in circulation and checking deposits at financial institutions and is considered the measure of money available for use in day-to-day financial transactions. Until the advent of NOW accounts a few years ago, the deposits in M1 did not earn interest.)

"Last year's decline may mark the beginning of a period of low or even negative M1 velocity growth as a result of paying interest on some fully checkable deposits . . . . It may have resulted from a one-time increase in money demand. It may have been a normal cyclical phenomenon," Weintraub said. "We don't know now which of these explanations is correct and cannot know for at least a year."

Weintraub, who believes strongly that targeting M1 growth is the proper approach to implementing monetary policy, hopes the unexpected drop in velocity was just a response to the recession. If not, he added, then some other target would have to be used because the policymakers at the Fed could never be sure again that the velocity of M1 and, thus, the expected rate of growth of GNP in current dollars might not turn out like last year--about 5 percentage points lower than intended.

Even a year may not tell the final story on velocity in 1982. There was a similar but smaller shift in the demand for money in 1975, and economists are still debating what happened and why.

But the Fed cannot wait a year or even a month to see what is happening to the velocity of M1 or M2. (M2 is a broader measure that also includes the new money-market deposit accounts, other savings and small time deposits, money-market mutual fund shares, overnight repurchase agreements and some other items.) It must have a policy directive at all times in the hands of its open market desk in New York, and on a daily basis it must decide whether to intervene in financial markets by buying or selling government securities. Buying securities adds cash, or reserves, to the banking system, while selling securities has the opposite effect.

Since last October, the Fed has had no short-term target for M1 growth because of the distortions expected from the regulatory changes. Instead, it has been focusing on M2 and M3, a still broader aggregate that also includes large time deposits, such as $100,000 certificates of deposit.

But now the rush of funds into the MMDA's is swelling M2, while greatly slowing the increase of M3. Volcker last week estimated that the January increase in M2 would be in the neighborhood of $50 billion, a gain that, if sustained, would involve a rise of 30 percent or more for the year. The M2 growth target tentatively set last July for 1983 was a range of 6 percent to 9 percent.

In the midst of all this, the Fed is trying to muddle through. It continues to set a path for the course of total reserves for the banking system and, for day-to-day operations, a path for non-borrowed reserves. In practical terms, the operations have assumed a flavor of an approach the Fed used quite a few years ago of focusing on the level of net borrowed reserves, according to officials there.

Some institutions end up with more reserves than they are required to have in a given week, while others have to borrow directly from the Fed. Net borrowed reserves, a figure the Fed doesn't even publish explicitly, is borrowed reserves minus any excess reserves for the banking system as a whole.

Such an approach of necessity relies more on judgments about what is going on in financial markets than on hitting targets for money growth.

"We . . . have had to approach monetary targeting and our operational decisions to provide reserves with greater elements of judgment and flexibility in the light of emerging developments," Volcker said in his JEC testimony.

That is exactly what has some conservative economists, including many of the monetarists, worried. They are afraid that when the Fed turns to judgment instead of a firm rule for setting policy that it inevitably ends up being inflationary.

William Fellner and Rudy Penner of the American Enterprise Institute last week expressed their concerns that the Reagan administration and, perhaps, the Fed would be aiming for too high a growth rate of nominal GNP--that is, GNP in current dollars--and thus set their monetary sights too high.

The Fed has "left everyone in complete uncertainty about how much money they want to pump into the system," Fellner said. "Now they have the freedom to do whatever they think is wise . . . . I'm quite pessimistic about that."

Penner added that under current circumstances the Fed cannot be held "accountable since there is no standard against which they can be judged." If the aggregates are not an adequate guide for policy, then the Fed should be asked to specify how fast it wants nominal GNP to grow, he said.

For its part, the Reagan administration shows nominal GNP rising 8.8 percent this year, more than 9 percent next year and then gradually declining to 8.6 percent in 1988. One administration economist said such a path was chosen because, "There is no analog in history for a period of sustained real growth and substantial declines in inflation."

To Fellner and Penner, that course for nominal GNP is too high to be consistent with continued declines in inflation.

Actually, with the recovery from the longest postwar recession just beginning, the administration may be willing to aim even higher. Last week Martin S. Feldstein, chairman of the Council of Economic Advisers, said approvingly that if the recovery began in early January, real GNP growth this year would reach 5 percent.

If it does, will inflation, as shown by the GNP deflator, fall to about 3.8 percent? Or will a larger increase in nominal GNP be tolerated so as to accommodate the 5.6 percent inflation rate forecast by the administration as well as the 5 percent real growth? Of course, the recovery may prove to be every bit as modest as shown by the official forecast, 3.1 percent rather than 5 percent real growth.

Either way, the Federal Reserve, far more than the administration, will determine the outcome. The FOMC this week probably will do little to change the tentative targets for 1983 it set last summer, though some type of allowance has to be made for the chaos caused by the regulatory changes.

The Fed cannot do what a number of officials there would prefer; namely, not set any targets until the situation stabilizes. The law requires they be set and reported.

But whatever targets do emerge this month, they will be tentative and will do little to end the confusion in financial markets. They also will not reassure critics such as Olsen and Fellner. In the weeks ahead, the Fed will continue to target on judgment.