I wrote the attached to help me decide whether to postpone or abort a hike due to lightning/heavy rain risk and it may be of interest to weather nuts. Following the detailed discussion is an abbreviated process one could carry in the field. Seasoned outdoors people may not find anything new here but I feel more confident in assessing risk after using this more methodical approach. Excuse the goofy formatting in trying to translate from the original word document, which the site wouldn't let me attach.

Western Colorado Thunderstorms

Disclaimer
Avoiding lightning in the backcountry is an individual responsibility that no guideline such as this can replace. Formation and intensity of mountain thunderstorms depend on multiple variables that even the forecasters don’t fully understand. Outdoor activities always involve risk and each person needs to understand their level of risk aversion. I wrote this guide to supplement anecdotal knowledge and to help me decide whether to postpone or abort a hike. Limitations include lack of statistical analysis to verify the data is robust enough to reach my conclusions and possibly biased data since it only covers 1-2 years.

Scope
My focus is on the western third of Colorado but it probably is useful for the central mountains also. This document doesn’t apply to the front range since weather patterns there are different. Note the following counties are designated hot spots for Colorado lightning by reference 1:
• Eastern Clear Creek
• Southern & far western Jefferson
• Douglas
• Western Elbert
• Teller
• El Paso
• Far northeast Fremont
• Southern & far western Las Animas

Detailed thunderstorm risk checklist (abbreviated version follows)
For those not particularly interested in weather forecasting, important items are marked with “*”, skip the rest.

1)Mountain hiking in summer always involves lightning risk. Determine your risk tolerance before the trip & use this checklist as a guide to supplement your decisions and not as a process to follow blindly.

2) Before a trip, verify all the links below are in your cellphone & display properly in your browser

3) Obviously forecasts issued closest to the time you hike are more accurate, especially if the forecast discussion states that models disagree with each other. Forecasts more than 3 days out are particularly susceptible to error.

*4) Local National Weather Service (NWS) office home page (e.g. http://www.weather.gov/gjt/)
a) Verify the local office is responsible for your counties of interest by scrolling to the bottom & clicking “our office” under “additional information”, then “forecast area maps”.
*b) *On the home page, check the map for colored areas indicating watches/warnings. Click the link to the right corresponding to the warning color for details. If a flash flood watch is in effect, I cancel canyon hikes or ones with major river crossings. For severe thunderstorm watches I cancel all hikes. For wind advisories or warnings I cancel high elevation hikes.
*c) Click the map at the location of interest to open the forecast page. Scroll down to a map that is to the right of “Detailed Forecast”. Click at a location that represents the highest elevation of your hike. Below the resulting map, verify the elevation is correct. The detailed forecast is now customized for that elevation. Note the percent chance of storms & what time they are expected to begin. Below the “Detailed Forecast” click “Tabular Forecast” to find hourly predictions of wind speed, temperature, precipitation potential & lightning chances. Based on analysis of a specific location, the time thunderstorms are expected to first appear seems to represent the earliest most, but not all, lightning strikes will occur so it is possible but not likely to have strikes before that time. I analyzed 45 days of forecast data from 7/7 to 9/15/17 for a 10k elevation location (Grand Mesa near Grand Junction) & compared it to actual storm coverage for west central Colorado, dividing coverage into categories of low/none, moderate and high, based on a purely subjective look at lightning maps. Thunderstorm predictions of 20% or less almost always resulted in no or low storm coverage. Half the time, 30% prediction resulted in moderate coverage. 40-60% resulted in moderate coverage 1/2 the time and high coverage 1/4 to 1/3 the time. >70% always resulted in high coverage. Admittedly this is a small sample size but here are the guidelines I came up with

0-20%: good chance of no storms but cannot rule them out
30-60%: assume storms can occur
70-100%: consider canceling the hike

*d) Click “Forecast Discussion” below the map. These are issued twice/day, usually between 2-4:30 AM and 2-4:30 PM with updates in between if warranted. This gives you an idea how confident the forecaster is. Note comments about model uncertainty, which reduces forecast accuracy.
*e) Below & to the right of “Detailed Forecast”, click on the radar image. The image has few landmarks so know how to find your hike location before the trip. Below the radar image, check all the boxes. To the left under “Short Range Images, click “Composite-Loop”. Key items to look for
--Speed/direction of storms. Remember that storms can intensify, collapse & intensify again. They can also collapse & trigger a new storm in a nearby location. They can form right over you, in which case radar is useless. Slow storm movement means prolonged rain and lightning.
--Training (one storm closely following the track of another) indicates flash flood risk
--Areas that are green and below usually lack lightning but rapidly intensifying areas are concerning no matter what color since mountain storms can progress from nothing to severe within 30 minutes.
--For canyon country, know the extent of the watershed for your hike. A severe storm some distance away can cause flash flooding.
f) Note I didn’t attempt to investigate mobile radar apps

5) Storm frequency. I divided Colorado into 3 parts from north to south & 3 from east to west, not including the front range since weather patterns there are different. Clockwise from the northwest corner the resulting regions are northwest, north central, mid central, south central, southwest and west central. For all days between 6/6 & 9/15 for 2016 & 2017, I visually assessed the lightning maps for how much of each region experienced lightning that day.
a) Results. Typically 30-50% of summer days will have moderate to high lightning coverage. Details below show year to year variation can be high. Reference 1 confirms that the San Juans have the 2nd highest lightning activity in Colorado with areas of the Front Range rated highest.

% days with no or almost no lightning. These days tended to occur consecutively, linked to persistent weather patterns.
June: 40-55% in 2016 with south central the least, 70-85% in 2017 with north central the least.
July: 50-60% in 2016, 5-25% in 2017 with southwest & south central the least
Aug & 1st half Sept: 40-60% in 2016 with southwest & south central the least, 30-55% in 2017 with southwest & south central the least

% days with moderate to high storm coverage
June: 35-50% in 2016 with southwest, south central & north central the highest, 10-15% in 2017
July: About 30% in 2016, 30-85% in 2017 with southwest & south central the highest
Aug & 1st half Sept: 30-50% in 2016 with southwest & south central the highest, 35-50% in 2017 with mid-central the highest

*6) Time of first strike: Reference 1 shows strikes before 10AM are rare. An uptick in average annual lightning activity occurs between 10 & 11 AM for a few widely scattered hotspots with more widespread activity starting between 11AM & Noon.

7) Precipitable Water: Go to this link: http://weather.unisys.com/current-weath ... nv=0&t=cur (under “Source” click “Upper Air Data”, under “Channel” click “Precipitable Water”) and verify it is current looking at the date/time above the map. These are issued for 6AM MDT (12Z) & 6PM MDT (00Z of the following day). Contour boundaries are labeled in 0.1” increments. Interpolate for your desired location. Analysis of 2017 storm coverage data indicate 0.7” or less correlates with low storm coverage, 0.8 & 0.9” are indeterminate & 1.0” or greater correlate with high coverage. This especially applies to southwest & south central Colorado, where monsoonal moisture surges are more common.

8) Current lightning maps: I haven’t checked into mobile apps so will not address those. One desktop site is http://en.blitzortung.org/live_lightnin ... php?map=30 Choose “North America” at left. To zoom into an area, at page top choose “Real Time Maps”, then “Dynamic Size”, pan to desired area & zoom. Recent strikes are white & change color as they age. Click the settings icon upper left to see strike age.

9) Morning cloud observations were made to see what relationships they have to later storm formation, defined as the occurrence of lightning by 2:00PM. This is for storms forming above 7-8000’ terrain in the Grand Junction area so comments may or may not apply to higher elevations.
--Clear skies at 10:00AM cannot predict risk of storms by 2:00 PM
--On days with lightning by 2:00 PM, building cumulus (obviously taller than wide) will typically appear by 11:00AM. But building cumulus doesn’t guarantee storms will occur.
--Lack of any type of cumulus by 11:30AM indicates delayed (after 2:00 PM) or no storm formation will occur.

Abbreviated thunderstorm risk checklist
1) Local NWS office home page
a) Check for watches/warnings on the home page
b) Click the map at the location of interest to open the forecast page. Scroll down to a map that is to the right of “Detailed Forecast”. Click at a location that represents the highest elevation of your hike. Below the resulting map, verify the elevation is correct. The detailed forecast is now customized for that elevation. Note the percent chance of storms & what time they are expected to begin.
--0-20%: good chance of no storms but cannot rule them out
--30-60%: make the assumption storms can occur
--70-100%: consider canceling the hike
c) Below & to the right of “Detailed Forecast”, click on the radar image. To the left under “Short Range Images, click “Composite-Loop”.
d) Click “Forecast Discussion” below the map. Note comments about model uncertainty, which reduces forecast accuracy.

2) Click “Forecast Discussion” below the map. These are issued twice/day, usually between 2-4:30 AM and 2-4:30 PM with updates in between if warranted. This gives you an idea how confident the forecaster is. Note comments about model uncertainty, which reduces forecast accuracy.

3) Time of first strike for 2017. Strikes before 10AM are rare. An uptick occurs between 10 & 11 AM for a few widely scattered hotspots with more widespread activity starting between 11AM & Noon.

4) For current lightning maps go to http://en.blitzortung.org/live_lightnin ... php?map=30 & at left choose “North America”. To zoom into an area, at page top choose “Real Time Maps”, then “Dynamic Size”, pan to desired area & zoom. Recent strikes are white & change color as they age. Click the settings icon upper left to see strike age.

5) Cloud observations. Relation of morning observations to the occurrence of lightning by 2:00PM.
---Clear skies at 10:00AM cannot predict risk of storms
---On days with lightning by 2:00 PM, building cumulus (obviously taller than wide) will typically appear by 11:00AM. But building cumulus doesn’t guarantee storms will occur.
---Lack of any type of cumulus by 11:30AM indicates delayed (after 2:00 PM) or no storm formation will occur.

References
1: Hodanish, Stephen & Wolyn, Paul: Lightning Climatology for the State of Colorado. NOAA/NWS Pueblo, Colorado. Internet document: https://www.weather.gov/pub/LightningClimatology
2: Vogt, Brandon J., Hodanish, Stephen J.: A High Resolution Lightning Map of the State of Colorado, 2014. Internet document: https://www.weather.gov/pub/LightningClimatology
3: Lightningmaps.org provides lightning history. Click archive, choose usa (big) map, date, time in UTC & dropdown gives length of dataset in hours from that time forwards. So in summer where local time is UTC-5 hours, setting it to 12:00 + 6 hours will give you what happened between 0700-1300 local time. If you turn animation on, it will default to clock setting + 2 hours, change as desired. Above the map you can pause animation or single step with prev/next

Reference notes
Figure 2 in reference 1 shows the average annual lightning flash density for the state. The density categories are in flashes/year/square kilometer or flashes/year/0.6 square miles. Since lightning can strike 15 miles from a storm, the most conservative way to estimate average risk from this map is to calculate the flash density for a circular area with radius of 15 miles = pi * (radius squared) = 3.14(15*15) = 707 square miles or 1178 square kilometers. For example, the darkest green area of the map would have on average up to 3*1178=3534 strikes/year that are within 15 miles of a given location. Since strikes that far away from a storm are rare, you can recalculate according to your risk tolerance. Choosing a 5 mile (8.3km) radius shows the medium blue areas would have up to 216 strikes/year on average & since most strikes are from May-Sept that comes out to 1.4 strikes/day on average. It could be much higher since I found large year to year variations so one should not assume blue on this map means you can ignore lightning risk.