Ever planned your outdoor adventure around the "perfect weather" forecast only to be disappointed when the skies darken and the downpour ensues? Kind of feels like nature is letting out a big laugh at your expense. It makes you want to go down to that news station where you received this forecast of utopian weather and let them know just how much the forecaster needs to find another profession. Predicting the weather is not an exact science. Meteorologists understand this, thus we get those frustrating 50% chance of thundershower days. One thing anyone can do to improve weather forecasts, and seems to always be underestimated, is simply add to the number of official surface weather stations.
We have all seen the television forecasters go to their all mighty computer weather forecasting models to show us exactly when and where the precipitation is going to fall. It all looks very impressive as the front pushes through at precisely 4:30 PM and the skies are clear by 7:00 PM. These computer models are very valuable and are also, deservedly so, here to stay. One of the problems with these models is that they can only be as good as the data that is put into them. Data is gathered from different sources including satellites, radar, weather balloons and of course surface weather stations. These surface weather stations are very important because the data is being produced in-situ. This data is collected in the field at the site of the study thus making accuracy easier to be obtained and measured. Satellite and radar data are remotely sensed and particularly with using satellites for surface weather conditions can be subject to much interference.
The way this data is gathered and the way the information is processed in the models can get very complicated. One of the basic concerns that meteorologists address with these models are the horizontal and vertical resolutions. Vertical resolutions refer to the spacing of data and model forecasting computations as you go up in the sky vertically. Horizontal resolutions are the spacing's at any particular level of the sky including the surface. Increasing the amount of surface weather stations will increase both the accuracy and the resolution of this horizontal data on the earth's surface. There are other variables such as how powerful the computer processor and the basic way the model is programmed that effect the resolution of the actual computations which affect the forecast. No matter how good the program or how powerful the computer the weather forecast is still unnecessarily flawed when it starts with inaccurate data.
To emphasize the point of why we need more surface weather stations it is useful to point out how "spotty" a weather system can be. We all know that it can rain on one side of town or even on one side of the street and be bone dry on the other. Rain will obviously affect the surface humidity. If this moisture is not initialized in a computer model correctly it could have a drastic consequence on your local forecast. Another thing to be considered is that this rain that took place where there are no weather stations is not even being recorded for climate data. This can throw off averages and hinder weather forecasting research as it appears to a researcher that the particular atmospheric environment did not produce precipitation when it truly did. If you want more information on how you can set up your own weather station that transfers data directly to the National Weather Service check out this guide from the Citizen Weather Observer Program . If you would like to venture into the challenging world of weather here is my article on how to become a meteorologist. Just know you may be held responsible for "bad" weather!