Weather Forecast Project ATMO 336

Case Study on the Accuracy of 500 mb Height Forecasts from the GFS Forecast Model 

Author: Emma Gillette  

1. Introduction 

This case study will serve to examine accuracy for 3, 5, 7, 10, and 15-day forecasts from the American GFS model. We will do this by comparing forecasts of 500 mb height with =the actual height patterns over the area of the continental United States. The forecasts were started at 12Z on Sunday, January 21st. The 3-day map became valid at 12Z January 24th, 5-day at 12Z January 26th, 7-day at 12Z January 28th, 10-day at 12Z January 31st, and 15-day at 12Z February 5th.  

The weather forecasts to be studied are produced by computer-generated numerical weather prediction models. These models produce 500 mb pattern maps, which helps one easily view and visualize large-scale weather patterns. In order to produce these maps, the current height above sea level where the air pressure reaches 500 mb is measured simultaneously and compiled into one map. From there, these forecast models predict the future pattern of 500 mb heights. 

2. Discussion of the 500 mb height forecasts 

This section describes the GFS model weather forecasts of the 500 mb height pattern for 3, 5, 7, 10, and 15 days into the future over the continental United States. Section 2a provides a general description of the 500 mb height pattern over and near the continental United States. In section 2b, specific forecasts for Tucson, AZ and Cleveland, OH are made based on the forecasted 500 mb height patterns.  

2a. Description of the forecasted 500 mb height patterns over the continental US    

Day 3 Forecast.  

 There is a strong closed low and trough just off the northwest coast with moderately below average heights for coastal regions and a chance for precipitation. There is a ridge covering much of the rest of the western United States with near average heights close to the Canadian border and slightly to moderately above average heights elsewhere. There is a broad trough across the eastern half of the country. The 500 mb heights are near to slightly below average in the northern part of the trough and slightly above average across the far south. There is a shortwave trough and ridge on the western edge of the broad trough over the states of Indiana and Ohio. 

Day 5 Forecast.  

There is a broad trough beginning at the continental west coast and ending in the center of the U.S., with moderately below average heights beginning north of Arizona and New Mexico. Along the coast of California and the Arizona-Mexico border heights are average. Crossing over the Canadan border is a closed low. There is a broad ridge over the mid-eastern part of the U.S., with heights moderately above average, with the greatest above average heights in the great lakes region. 

Day 7 Forecast.  

Day 7 shows an amplified pattern. There is a weak closed high both off of southern California and Southern Florida. Within the Western region there is a trough with moderately above average heights. Over the central planes there is a trough with contours packed closely together in the south and further apart in the north. Along the East Coast heights are rising into another ridge, with heights slightly to moderately above average in the South, and more above average in the North.  

Day 10 Forecast.  

There is a moderately zonal pattern for day 10. There is a broad, weak trough in the north west. Heights are near average to average in the South West, and moderately below average in the North West. In the Eastern half of the country there is a broad, weak ridge. Heights are near average in the states surrounding the Gulf of Mexico and are a bit more than moderately above average surrounding the great lakes region.  

Day 15 Forecast.  

There is a strong trough over the Western region, closer to coastal states. Height patterns are most severely below average over Idaho and surrounding states including California. The center of the continent is average height at the contours begin to rise into a more zonal pattern out of the amplified pattern of the west. The North East coast has heights that are severely above average.  

2b. Forecasts for Tucson, AZ and Cleveland, OH based on forecasted 500 mb Height Pattern 

Forecast  Time	500 mb  Height  (meters)	Average  500 mb  Height	Expected Weather
Day 3	5740	5680	Moderately above average temperature since 500 mb height is 60 meters above average. No chance for rain with ridge overhead.
Day 5	5680	5680	Completely average temperature since 500 mb is same as average. No chance for rain with zonal pattern.
Day 7	5780	5680	Moderately above average temperature since 500 mb height is 100 meters above average. No chance for rain with ridge to the west.
Day 10	5720	5680	Slightly above average temperature since height is 40 meters. Rain possible with shortwaves plentiful.
Day 15	5550	5680	Moderately below average temperature since 500 mb height is 130 meters below average. High chance of rain because it is downwind from a trough.

Table 1. Specific forecasts for Tucson, AZ based on the 3, 5, 7, 10, and 15 day forecasts 

Forecast  Time	500 mb  Height  (meters)	Average  500 mb  Height	Expected Weather
Day 3	5410	5450	Slightly below average temperature since 500 mb height is 40 meters below average. Precipitation is possible given the broad trough, especially considering the shortwave just off to the west, although there is a weak shortwave ridge overhead.
Day 5	5640	5450	Moderately above average temperature since 500 mb height is 190 meters above average. Clear weather likely because it is downwind of a ridge.
Day 7	5520	5450	Moderately above average temperature since 500 mb is 102 meters above average. Precipitation possible with trough to the west.
Day 10	5650	5450	Slightly more than moderately above average temperatures with 500 mb height 200 meters above avereage. Clear weather likely since a ridge is above.
Day 15	5640	5450	More than moderately above average temperature since 500 mb height is 190 meters above average. No chance of rain since it is at the top of a ridge.

Table 2. Specific forecasts for Cleveland, OH based on the 3, 5, 7, 10, and 15 day forecasts 

3. Evaluation of the 500 mb height forecasts 

This section evaluates the accuracy of the GFS model weather forecasts for 3, 5, 7, 10, and 15 days into the future that were described in section 2. In section 3a the forecasted 500 mb height patterns are evaluated by comparing then with the actual 500 mb height patterns measured at the valid forecast time. Section 3b evaluates the accuracy of the specific forecasts for Tucson and Cleveland.   

3a. Evaluation of forecasted height patterns over the continental United States 

Day 3 Evaluation.  The forecast was quite good for the most part. There is a trough and closed low off the northwest coast, a ridge in the west and a broad trough in the eastern United States. The western ridge is slightly stronger (higher heights) than predicted. Within the eastern trough, the shortwave trough and ridge that was forecasted to be over the states of Indiana and Ohio was actually slightly stronger (deeper trough and higher ridge) and positioned slightly to the west of where it was predicted to be.    

Day 5 Evaluation.  

The forecast is mostly accurate, with temperatures more intense than predicted in the Northwest and less intense than predicted in the Northeast. There is a trough in the northern midwest with moderately lower than average temperatures than expected and a lower trough than expected. The ridge above the great lakes region is nearly identical than expected, with more moderately above average temperatures.  

Day 7 Evaluation.   

While predictions for the Western part of the continent were mostly accurate, the heights were slightly lower than expected in the center of the country. The trough in the verified map is, on average, 10 meters deeper and the temperature was slightly below average rather than slightly above average. There was also an unpredicted shortwave within the western costal states. The ridge along the eastern coast is slightly higher than expected. 

Day 10 Evaluation.   

Predictions are becoming less and less accurate. Rather than temperatures moderately below average in the Northwest  and nearly average temperatures in California and Arizona, in the northwest there are moderately above and below average temperatures and in California and Arizona they are moderately above average. In the great lakes region the temperatures are not moderately above average but nearly average.   

Day 15 Evaluation.   

Very inaccurate. Prediction map shows a trough in the Western region and a ridge in the Eastern region. Verify map shows a less amplified pattern with the end of a ridge and moderately above temperatures in the west and a weak trough along the east coast. 

3b. Evaluation of the Forecasts for Tucson, AZ and Cleveland, OH 

Forecast  Time	Forecast  500 mb  Height  (Table 1)	Observed  500 mb  Height	Evaluation of the Expected Weather
Day 3	5740	5740	500 mb height was right on. Temperature and precipitation forecast was very good.
Day 5	5680	5720	500 mb height 40 meters off. Temperature above predicted, precipitation forecast good.
Day 7	5780	5800	500 mb height 20 meters off. Temperature and rain forecast good.
Day 10	5720	5790	500 mb height 70 off. Slightly above temperature than predicted, rain forecast possibly accurate with microwaves.
Day 15	5550	5770	500 mb height 220 meters off. Temperatures much higher than predicted, no chance of rain.

Table 3. Evaluation of forecasts for Tucson, AZ 

Forecast  Time	Forecast  500 mb  Height  (Table 2)	Observed  500 mb  Height	Evaluation of the Expected Weather
Day 3	5410	5440	Actual height 30 m higher than predicted, which is near average, while forecast was slightly below average. Precipitation is less likely since shortwave trough is further west and Cleveland is located just downstream of a shortwave ridge.
Day 5	5640	5370	500 mb height 270 meters off. Weather and precipitation predictions good.
Day 7	5520	5550	500 mb 20 meters off. Predicted above temperatures correct and slight chance of rain predicted correct.
Day 10	6560	5520	500 mb height way off at 1040 meters below predicted height. Weather less above average than predicted, precipitation prediction good.
Day 15	5640	5370	500 mb height 270 meters off. Temperature much lower than predicted and precipitation prediction correct.

Table 4. Evaluation of forecasts for Cleveland, OH 

4. Discussion 

According to this particular study, the temperature predictions of forecasts at earlier days are more accurate, with predictions growing less and less accurate each day. 500 mb height predictions generally followed the same pattern, with the most dramatic inaccuracy leaving a 1040-meter difference on day 10 in Cleveland. Precipitation predictions seemed mostly accurate. Note that the results are based on a single case study and may not be representative of the forecast accuracies for the GFS model in general.  

Explain why water will boil at a lower temperature on top of a 12,000 foot (3650 meter) mountain?  

At higher altitudes water boils at a lower temperature because air pressure is lower. Water boils when the vapor pressure equals the air pressure of the atmosphere, and since it is lower at higher altitudes, the water will boil with less energy (less temperature).  

Suppose you are camping at 12,000 ft (3650 m) altitude and cooking food with boiling water. What adjustments must you make compared to cooking with boiling water at lower altitude? 

You would have to cook food a little longer. This seems backwards since the water boils faster, but actually the water would be cooler than boiling at sea level.  

Which of these may help to increase the temperature of the boiling water: (a) Simply "turn up the heat" on your camping stove or (b) Put a heavy lid on top of the pot of boiling water? 

b. It reduces the vapor that can escape, and like the figure depicted in the reading, it will increase the vapor pressure in a closed system. This causes the air to become saturated in the pot, and the temperature may become warmer.  

Because water boils when the saturation vapor pressure is equal to the surrounding air pressure, water will boil at a lower temperature at higher altitudes due to the surrounding air pressure being lower. This can be a problem at higher altitudes, like a 12,000 foot mountain, because when you cook in boiling water, it will not be as hot as it would be at sea level. To compensate, you either need to cook longer or use a heavy lid on the pot. A heavy lid works because it creates a closed system, allowing for the trapped air to reach a greater air pressure, therefore increasing the boiling point of the water. Simply turning up the heat will not work because once water reaches its boiling point, it will only turn into water vapor, and the liquid water will remain at the temperature of the boiling point. 

1. Since water boiling means the water has reached the saturation vapor pressure of the surrounding air, then when the surrounding air is lower pressures it is easier for the water to reach that pressure. Then the water will boil at lower temperature because it can quickly reach that required pressure at a lower temperature. We also know that high altitude mountains have lower air pressure, so high altitude mountains make the water boil at lower temperature because of the lower pressure. The first way will not work because turning up the heat does not change the pressure unless the gas can be in a fix volume, but since it is open, more heat will only speed up the rate of heating but not the boiling temperature. The second way effectively keeps the gas in a constant volume, so as it heats it increases in pressure. This higher pressure will force a higher saturation vapor pressure from the water and it will boil higher temperatures. 

 Water boils when the saturation vapor pressure, the extent of liquid molecules escaping into the vapor, equals the outside air pressure. When the air pressure decreases, such at a higher elevation, it takes less heat to match the saturation vapor pressure with the air pressure, leading to water boiling at a lower temperature. 

Food being cooked at high altitudes with boiling water would need to be cooked longer than at lower altitudes because boiling water is at a cooler temperature at high altitudes. 

b) because it would be a closed system and the air inside the pot would become saturated, aiding in high temperatures. 

Water boils at a low temperature at the top of a mountain because the air pressure is lower which makes it easier to reach the saturation vapor pressure. When cooking this means you must make adjustments because while the water boils faster, it boils at lower temperatures so you must cook food for longer. Putting a heavy lid on top of the boiling water would be best because it doesn't allow vapor to leave the pot which increases the vapor pressure, thus raising the temperature of the weather.