Depends what you mean by forecasting.
It's relatively easy to forecast long-range broad scale regional trends based on known factors that only change across significant timescales (ocean and air currents, solar activity, apogee/perigee, axial wobble). So, saying something like "this summer will be hotter and wetter than usual thanks to a prevailing El Nino" or "this winter will be milder than usual" is not particularly difficult.
It's much more difficult to make medium range (10-60 day) forecasts, and more difficult still to make those location specific. Meteorologists can look at local air pressure differentials, soil moisture content, air moisture content, sea temperatures and a couple dozen other metrics and come up with forecasts that are pretty good when there are wide scale patterns, but not so great for any specific area. Statements like: "There is a higher than usual chance of rain for Southern Florida over the next two weeks", or "the second half of August is likely to be hot and mostly dry across the Pacific Northwest" are going to be pretty good, but it's unlikely that a specific forecast like "it will rain in the morning in Galveston on Tuesday week" is going to be accurate.
It's really difficult to make intermediate range and location specific forecasts (2-9 days), unless there is a major set of prevailing atmospheric conditions dominating. If a big low pressure front is moving in, then the forecast is likely going to include a high chance of rain, high winds and lower temperatures. If a big high pressure region is prevailing, then conditions are likely to be dry and calm. However, if the is no strong atmospheric phenomenon, the picture is a lot less certain. And the further you drill down to a specific location, the less certain the forecast will be.
Short term forecasting (6-48 hours) becomes much easier, particularly with location specific forecasts. Because weather is necessarily local and there are good historical records about weather patterns available, forecasters can look at what the weather is doing in the area and in the neighbouring areas, along with wider scale trends, look at historical patterns and make reasonably accurate forecasts. So, a forecast like "Sydney will have cool temperatures and intermittent showers this morning, with the weather clearing by early afternoon" or "Perth will have a cold morning, with a late change leading to overnight rain" will likely be highly accurate.