Generally, weather-related science fair projects score well with teachers and judges because they require time and effort, much like plant projects. It is possible to do idea #2 or #3 in a weekend if the weather cooperates (i.e. it rains or snows when you need it to), but typically you will need at least 1 month, and 6 to 12 months are more common. So, if you haven’t started this year’s project yet, it might be too late, but this would be a great time to start next year’s project.
IDEA #1: One strategy for a weather-related project is to build your own piece of equipment (Part I) and then use that equipment to track the weather over a long period of time (Part II). For example, you could build a barometer from common household items and then compare its measurements to a “real” barometer, or the nightly news reports for atmospheric pressure. This would also work with a home-made thermometer or anemometer.
IDEA #2: Another idea would be to build your own rain gauges and use them to record rainfall relative to a spatial gradient. Perhaps distance from a building, stream, patch of trees, or other structure of interest. Alternatively, you could place them in cardinal directions around the structure or in areas of varying land use (e.g. rural to suburban to city). Add depth to this project by testing the rain water collected with a pH test kit from a pet store. Ultimately that would give you several variables to analyze (time, spatial gradient, rainfall totals, and pH).
IDEA #3: If you have less time, but live in a snow-prone zone, you could monitor snowfall amounts relative to spatial gradients within the local landscape. Or select items that might be used to pre-treat roads (salt, sand, gravel, kitty litter, etc) and put them in measured patches on your driveway/yard before a predicted storm. If you have access to a camera, set one up (from the window) and see if you can detect differences in snowmelt as snow falls on the different items (will also make great visual aids for backboard). At specified times, go out and measure the depth or percent coverage of snow on the patches. Don’t forget to replicate the patches.