September 22, 2022 Death Valley Too Hot Again

A recent blog entry by me which compared the official Death Valley temperature-measurement instrumentation with adjacent temporary equipment (installed by Dirk Baker of Campbell Scientific) showed that the official readings at Death Valley/Furnace Creek (DV/FC) on typical hot and sunny summer afternoons tend to drift too high, especially during light wind conditions. This was apparent on July 9, 2021, when the official maximum of 130F at Death Valley/Furnace Creek was not supported by the adjacent (temporary) sensors. The discrepancy appears to be due to a number of things:

–) The DV/FC weather station is above a natural (bare and dry) desert ground surface which heats up tremendously under the summer sun.

–) The DV/FC weather station is somewhat wind-protected by some nearby structures and vegetation to the south. Prevailing wind on summer afternoons here has a southerly component.

–) The current electronic instrumentation does not incorporate an aspirated shield, and the adjacent hardware and solar panel are very close to the thermometer shield and might contribute to higher (i.e., spurious) maximum temperature measurement.

Prevailing wind is from the south on summer afternoons at Furnace Creek, and an increase in structures and vegetation to the south of the DV/FC station in recent decades has resulted in an increase in average and extreme maximum temperatures. (see Roof and Callagan, 2003 BAMS) Lighter winds through the station area permit air heated along the ground to be less effectively mixed away. And, lighter winds exacerbate the equipment issues, with temperature drifting upward in the passive (non-aspirated) shield. The 30-year “normal” July average maximum at DV/FC from 1961-1990 was 114.3F, and from 1991-2020 was 117.4F. And since you asked, the normal for June increased by 3.0 degrees (F) and the normal for August increased by 2.8 degrees (F).

Yes, the hotter summer maximums in the more recent 30-year timeframe are due in part to hotter summer weather overall in the region. But, I believe that a significant chunk of the three-degree (F) increase at DV/FC is due to poorer ventilation through the station site area now versus 30-60 years ago. At Daggett AP, the increase in average maximum for July from 1961-90 to 1991-2020 was 1.1 degrees (F). At Bishop AP, the increase was a healthy 2.5 degrees (F), and at Tonopah AP the increase for July was 1.2 degrees (F).

The reason for this entry is to highlight one particular day this September which well-demonstrates this tendency for DV/FC temperature to drift too high on sunny summer afternoons with light wind. Here are the maximum temperatures in Death Valley on 9/22/2022, a day with wall-to-wall sunshine:

The official DV/FC station was five degrees warmer than Stovepipe Wells and four degrees (F) warmer than Badwater. Most of this temperature difference is simply due to the differences in station exposure, and the idiosyncrasies of the instrumentation. The DV/FC station is not nearly as well-exposed as the other four stations. The Cow Creek/CARB station was 9 degrees (F) cooler on high temperature compared to DV/FC. This (California Air Resources Board) station is along a well-exposed slope, and the temperature sensor is well above the ground (maybe 10 meters or more?), and thus maximums here are generally very conservative and at least a couple of degrees cooler than they would be compared to a standard station with the sensor about 5-6 feet above the ground. The Cow Creek station is only a few miles from DV/FC, and 600 feet higher in elevation. The elevation difference provides the DV/FC station with a 3-degree (F) advantage over Cow Creek/CARB on summer maximum temperatures. High temperatures at Cow Creek/CARB are routinely 6-7 degrees cooler than those at DV/FC in summer due to the exposure and elevation differences between the two stations.

Let’s take a look at some hourly data from September 22:

Death Valley/Furnace Creek below (temp/dew point/RH/heat index/wind dir/wind speed/prec)

Sep 22, 8:00 pm	94	32	11	89	S	1G3	0.00
Sep 22, 7:00 pm	95	32	11	89		0G0	0.00
Sep 22, 6:00 pm	97	36	12	92		0G0	0.00
Sep 22, 5:00 pm	97	36	12	92		0G0	0.00
Sep 22, 4:00 pm	96	38	13	92		0G4	0.00
Sep 22, 3:00 pm	95	37	13	90	WNW	2G6	0.00
Sep 22, 2:00 pm	94	38	14	90	WSW	4G7	0.00
Sep 22, 1:00 pm	91	39	16	87	S	3G5	0.00
Sep 22, 12:00 pm	89	39	17	85		0G3	0.00

Death Valley NP/Cow Creek/CARB

Sep 22, 8:00 pm	91
Sep 22, 7:00 pm	92
Sep 22, 6:00 pm	93
Sep 22, 5:00 pm	93
Sep 22, 4:00 pm	92
Sep 22, 3:00 pm	91
Sep 22, 2:00 pm	90
Sep 22, 1:00 pm	88
Sep 22, 12:00 pm	85

Saratoga Spring (temp/dew point/RH/wind/sunshine/precip)

8:34	87	30	13	83	NNW	1G3	0	—	1.37	0.00
7:34	93	24	8	87		0G2	0	—	1.37	0.00
6:34	97	23	7	90	WNW	2G8	222	100	1.37	0.00	0.00	0.00
5:34	97	27	8	91	NW	3G9	427	87	1.37	0.00
4:34	96	26	8	90	NW	3G9	607	91	1.37	0.00
3:34	95	30	10	89	NW	4G9	742	93	1.37	0.00	0.00
2:34	92	28	10	87	W	3G11	822	89	1.37	0.00
1:34	91	34	13	87	SW	3G10	843	85	1.37	0.00
12:34	89	34	14	85	SSE	3G9	802	82	1.37	0.00	0.00	0.00	0.00
11:34	86	36	17	83	SSE	3G10	699	79	1.37	0.00

Stovepipe Wells 1SW

Sep 22, 8:00 pm	91	22	8			G6	0.00
Sep 22, 7:00 pm	93	21	7			G8	0.00
Sep 22, 6:00 pm	95	22	7			G8	0.00
Sep 22, 5:00 pm	96	23	7			G9	0.00
Sep 22, 4:00 pm	96	26	8			G11	0.00
Sep 22, 3:00 pm	94	21	7			G9	0.00
Sep 22, 2:00 pm	93	29	10			G10	0.00
Sep 22, 1:00 pm	92	28	10			G10	0.00
Sep 22, 12:00 pm	90	32	13			G12	0.00

And for your convenience, here are the hourly temperatures for 3, 4, 5 and 6 p.m. PDT:

The hourly temperature data show a 4-5 degree spike in temperature at DV/Furnace Creek at 4 p.m. PDT, while each of the other stations changed by a couple of degrees from 3 to 5 p.m. PDT.

Wind on this afternoon was less active compared to most summer days. The average wind speeds provided on the hourly observations at DV/FC were only 1-3 mph from 3 to 6 p.m. PDT, with a peak wind gust of 7 mph. At 4 p.m., when the temperature was 102F at DV/FC, the wind direction was SSE (SSE2G6). There was a westerly wind component on the obs for 3 and 5 p.m., which were 97F and 98F, respectively. There MIGHT be a tendency for these afternoon temperature spikes to be more pronounced when light wind has an easterly component, as occurred on July 9, 2021, when the station reported 130F. Could it be that the adjacent solar panel and the equipment support hardware (mere inches from the radiation shield on its east side) is causing some spurious increase in temperature values?

Wind data at the much-better exposed Cow Creek/CARB station were missing. Hourly temperature differences with this station and DV/FC were a very reasonable 5-6 degrees (F) at 3, 5 and 6 p.m. PDT. The 5-6 degree differences can be attributed to the elevation difference (3 degrees F), station exposure differences, and sensor height differences. See my other blogs on the Death Valley station exposure issues for full explanations! Also keep in mind that the CARB station hourly temperatures are the highest readings attained during the previous hour. But, given the high sensor height, temperature changes are small and subtle at the CARB station hour-to-hour on summer afternoons comparably. And, one more thing to keep in mind on the Cow Creek/CARB station: its position on the west-facing slopes on the east side of the Death Valley basin may promote a slightly higher temperature there on mid-late summer afternoons (after the elevation part is factored out). Yes, there are plenty of influences on temperature in the desert on summer days. At least the ground cover is the same (barren) beneath all of these Death Valley stations. Hooray! So, now that all of this preamble has been presented, I can say that the 10-degree difference between DV/FC and Cow Creek/CARB at 4 p.m. PDT is not reasonable. The stations are only separated by a few miles, and DV/FC is only 600 feet lower. Cow Creek shows a smooth and slow upward trend in temperature from 3 to 6 p.m. PDT (as do Badwater, Stovepipe Wells and Saratoga Spring), while DV/FC has that prominent spike. The DV/FC temperature at 4 p.m. PDT would be considered somewhat questionable here.

The Badwater data show light winds at 2 and 3 p.m. PDT, and calm winds for the remainder of the afternoon and into the evening. It is quite uncommon to see such light winds at Badwater, and I have a hunch that the wind data on the hours showing gusts of zero mph are flawed. The Badwater station is very well-exposed in that there is little nearby to slow the wind (i.e., no vegetation or structures). The locality is in a bit of a wind-protected spot due to the cliffs to the east and a small ridge to the south.

Badwater, like Cow Creek, shows a slow and steady climb from 95F to 97F from 3 to 6 p.m. PDT.

Below are the hourly HADS data sets for DV/FC (on the left) and Badwater (on the right).

The 4th column (among those in white on the shows the peak wind speed (mph) for each hour (Z time). The next three columns provide the hourly temperature, the lowest temperature during the hour, and the highest temperature during the hour. The last two columns are RH and precipitation.

Before our temperature comparison between Badwater and DV/FC on this date, let’s take a look back to 1959-1961, when a station was established at Badwater during the summer months. The cotton-region thermometer shelter used at Badwater was the same as that at Cow Creek and Greenland Ranch (1959-1960) and at Death Valley/Furnace Creek (1961). Summer maximums at Badwater averaged a couple of degrees (F) hotter compared to Cow Creek and about 3 degrees (F) hotter than at Greenland Ranch. This was and is no surprise, as it has been known that summer afternoon temperatures tend to be a little higher at Badwater versus the Furnace Creek area. Badwater is about 100 feet lower in elevation, so that gives Badwater a 0.5F advantage versus Furnace Creek. The Furnace Creek stations (especially Greenland Ranch and DV/FC) were/are influenced some by the local irrigated areas and resulting evaporative cooling. And, Badwater is in a bit of a “notch” with respect to the local topography, with the steep west-facing slopes nearby, which may help some to promote localized hotter maximums compared to the more-open Furnace Creek area. (This of course does not consider the effects of wind-slowing structures and vegetation at Furnace Creek!)

My personal temperature measurements from Furnace Creek to Badwater and back on summer afternoons through the years supported the findings from 1959, 1960, and 1961 which showed that Badwater was some 2-3 degrees hotter than the Furnace Creek area stations.

A new automatic station was installed at Badwater in August 2021. It is a Campbell Scientific station and is quite similar to the official equipment at DV/FC. The radiation shield is NOT aspirated at Badwater (the same goes for DV/FC). The numbers for summer 2022 are just about in, and surprisingly (“surprisingly” if one is not up to speed on the DV/FC issues!) the maximums at Badwater have averaged slightly LOWER than those at DV/FC. Why is this? Well, you know what I think: The official DV/FC station tends to provide maximums which are some 2-3 degrees (F) too high because of ventilation and exposure issues, as airflow through the weather station area is decreased and heated air is not mixed out as readily. The lighter winds due to the obstructions also exacerbates the temperature measurement problem that is known to exist with its non-aspirated radiation shield and sensor. There is little reason (yet) to believe that the Badwater station maximums tend to trend too high. Exposure at Badwater is excellent and the general area is usually windier than the Furnace Creek area on summer afternoons. Note: The wind at Badwater goes calm for several hours around the hottest time on this (September 22) date, with even the peak gusts at 0 mph. The “look” of the wind data at Badwater during this week in September is a bit wonky, with plenty of “zero” readings. I think one could assume that the Badwater wind data are less than reliable, and that there was probably at least SOME light wind during the hours which show a peak gust of 0 mph.

The ranges of temperature by hour on September 22, 2022 were quite a bit larger at DV/FC as compared to Badwater:

Death Valley/Furnace Creek and Badwater, range of hourly temperatures (in F)

2-3 p.m. PDT FC 95.8 to 99.9 BW 93.1 to 95.2 (4.1 versus 2.1 degrees)

3-4 p.m. PDT FC 95.9 to 101.8 BW 94.4 to 96.4 (5.9 versus 2.0)

4-5 p.m. PDT FC 97.0 to 100.7 BW 95.5 to 96.9 (3.7 versus 1.4)

5-6 p.m. PDT FC 96.2 to 98.7 BW 96.0 to 97.7 (2.5 versus 1.7)

The main takeaways here:

–) DV/FC was hotter then Badwater by four to five degrees F (on highest temperatures during the hour) during the hottest hours of the afternoon! Badwater warmed a little from 5 to 6 p.m. PDT while DV/FC cooled a little. It is typical for Badwater to have rather late maximums due to its proximity to the steep and barren west-facing slopes to its east. Prior to 5 p.m., the day’s maximum temperature at Furnace Creek was 5.4 degrees (F) higher than at Badwater.

–) DV/FC was hotter than Badwater on all of the “lowest” hourly temperature values.

–) Hourly temperature range was roughly about 3-6 degrees F at DV/FC, and just 2 degrees F at Badwater. This difference is largely due to station exposure differences. The better-exposed station (Badwater) has better ventilation, and here there is much less of a proclivity towards a build-up of localized hot air around the instrumentation. The case is opposite at DV/FC: a poorer exposure, poorer ventilation, hot air builds from the ground and is less efficiently mixed out, ambient air temperature spikes some, then that air is replaced and a relatively large intra-hourly temperature range is the result.

I see little reason to do a detailed comparison between DV/FC and the Stovepipe Wells and Saratoga Spring stations! The highs of just 96F and 97F, respectively, at these two very well-exposed stations support the highs of 93F at Cow Creek/CARB and 98F at Badwater. These maximum temperatures do not support the 102F at DV/FC. Afternoon wind speeds generally reached 10 mph each hour during the mid-late afternoon hours at Stovepipe Wells and Saratoga Spring.

So to wrap up this little blurb…

The well-exposed low-elevation stations in Death Valley (Badwater, Stovepipe Wells 1SW, Saratoga Spring and the CARB station at Cow Creek) exhibited a typical and relatively smooth march of temperatures through the afternoon of September 22, 2022. The official NPS/NWS Death Valley station at Furnace Creek showed a rather pronounced spike in temperature at 4 p.m. PDT, allowing its daily maximum temperature to be several degrees hotter than at the other stations. A relatively wide range of temperature during the afternoon hours at DV/FC suggests that the station exposure and ventilation are at least somewhat inadequate. And, when winds in the general area are on the light side during the afternoon, the maximum temperature reports at DV/FC are inflated a bit more due the unaspirated radiation shield in use. This scenario was evident also on July 9, 2021, when the DV/FC station registered a maximum temperature of 130.0F.

I might add here that it is no secret that the unaspirated radiation shields are not perfect, and that measurement errors are expected with them when wind is very light and solar radiation is strong. It appears that the radiation shields in use at DV/FC and Badwater are the “Young” type. These are called “multi-plate radiation shields,” and temperature errors (due to radiation) of 2.7 degrees F are likely when wind is 2 mph or less! See these links for the Young radiation shield specifications!!

Young link

Young link

This Campbell Scientific web site provides an overview of the Young radiation shield that is used with its stations.

This link provides a couple of images of the new Campbell Scientific weather station at Badwater. It looks like the Young radiation shield is a bit higher than six feet above the ground, which would help a little to mitigate spurious upward temperature spikes during light wind on sunny afternoons (compared to shields mounted lower).

A big thanks to my buddy Dennis who has helped me with the instrumentation information.