Wednesday, March 14, 2018

March snow

The first week of March brought a net snow accumulation of nearly 50 cm to the Northern Icefield, which by any measure is a snowy interval on Kilimanjaro. This precipitation follows 25-30 cm of continuous ablation during February, as illustrated in the previous post. A context for the event follows.

Figure 1 (below) shows Sentinel-2 satellite images of the exact same scene, on the last day of February and on 5 March. As detailed in another post, snowcover was primarily confined to steep north-facing slopes by the end of February. Although considerable cloud cover is present around the mountain on the 28 Feb. image, the summit caldera is mostly cloud free. Note the red squares, which are co-located on the 5 March image for orientation. High clouds partially obscure the March image, yet pervasive snowcover is visible. A sharp snowline at ~4,400 m is visible on the left-hand side of the image.

Figure 2 provides two snowy views of the mountain from the Moshi area (SENE credit). Despite low resolution of the 3 March image (upper), substantial snowfall obviously occurred since the satellite image acquired 3 days earlier. Snowcover appears to be somewhat more uniform than it was on 8 March (lower) - consistent with the timing and magnitude of snowfall recorded at the summit weather station.

At the Northern Icefield, satellite telemetry (Argos) shows ~12 cm of accumulation on 2 March, ~15 cm on the 3rd, and ~5 cm on each of the next 4 days. The precision of these daily totals will be improved when higher temporal resolution data are recovered from the automated weather station. Due to the diurnal cycle of climate on the mountain, some ablation likely also occurred on most of these days and is probably responsible for the patchier snowcover on the 8 March image.

A fascinating element of this snowfall period is provided by a depiction of regional-scale circulation (Fig. 3; Cameron Beccario credit). Here, airflow on the morning of 4 March is illustrated at the 500 hPa pressure level, equivalent to Kilimanjaro summit elevation. Airflow at this level appears to have been influenced by Tropical Cyclone Dumazile beginning on the 2nd as the storm intensified, continuing through about 7 March. The relationship between Kilimanjaro snowfall and cyclones in the southwest Indian Ocean is being investigated with collaborators Thomas Mölg and Emily Collier (Friedrich-Alexander University), along with Timba Nimrod.

On this figure, Kilimanjaro's location is shown by the green circle. Note the westerly wind, which prevailed through the snowy interval. Wind measurements at the summit (via telemetry) verify this airflow, which is atypical at the summit (only ~5% of hourly means are from 270° ±30°). Riming of the instruments appears to have occurred during the event, causing data loss particularly on the 3rd, 4th, and 6th. Nonetheless, such verification of airflow by in situ measurements is not a trivial finding - for very few continuous meteorological measurements exist from nearly 6000 m with which to compare output from numerical models.

Finally, figure 4 depicts circulation and humidity on 3 March. Here the highest humidity is shown in cyan color, suggesting a Congo basin origin for this precipitation event.

Monday, February 26, 2018

NIF surface, mid-February

Here is the Northern Icefield surface on 15 February, courtesy of Thomas Lämmle (EXTREK-africa). Our two UMass weather stations are visible on either side of the guides, with Mt. Meru in the background.

This image is particularly useful in documenting the glacier surface. Beneath the 2 ultrasonic snow sensors the surface is uniformly flat, with minor penitentes resulting from ablation of January accumulation. Therefore, further changes in height recorded at the station should be nicely representative of accumulation/ablation changes over a larger area.

Since our fieldwork in early October, net lowering of the surface has been ~15 cm. The current glacier surface at the AWS appears to be comprised of transformed seasonal snow, which is considerably brighter (i.e., higher albedo) than the immediately underlying ice.

Further ablation and lowering of the glacier surface will be determined by when the long rains begin, which typically occurs early in March.

Sunday, February 25, 2018

Northern Icefield from Kenya

This is Kilimanjaro as viewed from the north this morning (Amboseli Park in Kenya, 8 AM on 25 February 2018). Part of the Northern Icefield is visible on the right-hand side of the summit. This northern portion, largely outside the crater (caldera) rim, has now separated from the southern part of the glacier which most climbers see from Uhuru Peak and within the crater.

Although the Northern Icefield is the largest glacier on the mountain, it is shrinking rapidly. The image below depicts the glacier about 20 years earlier; this is an aerial view looking south.

Thanks as always to Simon Mtuy for sending photos!

Saturday, February 24, 2018

Tuesday, February 20, 2018

January snowcover [updated]

The 2017 'short rains' season brought very little snow accumulation to the summit this year. Although snowfall is often variable during the short rains (typically November and December), this year was especially dry and resulted in a net lowering of glacier surfaces.

Early January brought the first noteworthy snow accumulation to the glaciers since the dry season began in June 2017. The graph above,
illustrating January snowfall, is based on snow measurements obtained from the Northern Icefield AWS by satellite telemetry. Such measurements must always be viewed cautiously, as wind redistribution of snow and other factors can complicate data interpretation - yet as average measurements from 2 sensors, their reliability is improved.

Snowfall is critically important to Kilimanjaro glaciers, primarily by controlling the amount of solar radiation reflected from the surface. When a bright snowcover exists, the reflectivity (albedo) is high; aging snow becomes gradually less reflective, and thinning snow allows radiation to penetrate through to the underlying glacier ice.

On the graph above, numbered squares correspond to a selection of natural-color images from the ESA (European Space Agency) Sentinel-2 satellite, shown below. These depict snow accumulation steadily increasing during early January, culminating in the events of 10 and 11 January. With little precipitation through the balance of the month and into early February, the mountain's snowcover then thinned and became patchier.

Here are a few notes on the Sentinel-2 images, which qualitatively demonstrate that
Kilimanjaro glaciers are sensitive to the magnitude, frequency, and spatial extent of precipitation events!

Image #1, 30 December 2017:  Clouds to the west of Kibo, filling the Western Breach and obscuring the southern slope glaciers. Within the caldera and south of the crater rim, uniformly-white areas are glaciers (likely with some snowcover). No snowcover is present within the caldera, with only a light dusting on the highest northern slopes, and some accumulation on the north-facing section of the caldera rim (e.g., below Uhuru Peak).
Image #2, 9 January 2018:  Extensive clouds, yet uniform snowcover is visible within the caldera. Note the lack of snow on a small portion of Reusch Crater (dark area), where geothermal heat flux is probably responsible for melting any accumulation. The graph above shows that this image represents snowcover prior to the two largest snowfall events of January, on the 10th and 11th.

Image #3, 14 January:  Little or thin cloud cover is present over the Kibo caldera, in contrast to the clouds obscuring the mountain's western and southern flanks. Snowcover is extensive, as suggested by the graph above.

Image #4, 24 January:  Scattered clouds surround the mountain at elevations below ~5,000 m. Snowcover from early January has thinned within the caldera and on the glacier (see graph), and generally become patchier. Albedo of the glacier surfaces remains high, reflecting most of the incoming solar radiation. The Reusch Crater is now mostly snow-free, and limited snow remains within the Western Breach. Climbers going to the summit on this date were likely walking on snow most of the distance above Stella Point.

Image #5, 8 February:  A typical February day on Kibo, with scattered clouds concentrated to the south and west. Snowcover patches within the caldera have shrunk further. Note the almost complete lack of snow on eastern and south-facing slopes; even at only 3° south latitude, solar radiation receipt is greater on south-facing than north-facing slopes in early February!

Image #6, 18 February (not shown on graph):  Without measurable snowfall since image #5 was acquired (see above), the extent of snowcover continues to decrease, lingering primarily on steeper north-facing slopes. During this time the Northern Icefield surface decreased in height by another 10 cm, and with ablation of January snow the surface is likely bare glacier ice once again. On the southern slope of this image, distinguishing between snow and ice is difficult for those unfamiliar with the glaciers, yet these ice bodies were all contiguous only ~10 years ago. As shown below - and explained in the 21 November post - this is no longer the case, as the glaciers continue shrinking.

Tuesday, November 21, 2017

South-side glaciers

As late as the 1980s, three distinct bodies of ice remained on Kilimanjaro's caldera rim and spilled over onto the slopes. A century earlier, these icefields - the Northern, Eastern, and Southern - were more-or-less connected as one large ice cap. The map below (after Hastenrath, 1984) depicts the three icefields along with their outlet glaciers, which became more distinct as the ice thinned and retreated.

The mountain's south side is shown in the image above, from a rarely-seen perspective above the village of Mweka. The Diamond Glacier can be seen on the left-hand side; this thin feature appears more like a perennial snowpatch than a glacier, supported by the recent appearance of rocks protruding through the snow and/or ice. At the center of the image is the Kersten Glacier, which separated into upper and lower sections ~10 years ago. Two small blocks of ice left of the Kersten are all that remain of the Heim Glacier, which rivaled the Kersten in length 20-25 years ago. The fragmented ice on the right-hand side is what remains of the Decken Glacier. And in the upper-right corner is the Rebmann Glacier; this portion of the former Southern Icefield is visible from the trail between Stella Point and Uhuru Peak - as well as from Barafu Camp.

Glaciers on Kilimanjaro remain beautiful, and continue to reveal secrets of both their history and that of the mountain's climate. The "Roof of Africa" will be a very different place when the glaciers are gone.

Friday, October 20, 2017

At the summit: days #92-96 in the crater

This month we returned to Kilimanjaro's summit glaciers and automated weather stations, 14 months since our last visit (Aug. 2016). Yes, the changes were dramatic - everywhere we looked.

This post provides a few glimpses of the remaining ice, still incredibly beautiful. Once an initial inspection is done on recovered AWS data, a subsequent post will provide an overview.

Helping out on this fieldwork were Spencer and Chang'a (Fig. 1). This was both of their first times on the mountain and both brought new insights and questions, providing stimulating discussions during the ascent and in camps. Dr. Ladislaus Chang'a is Director of Research and Applied Meteorology at the Tanzania Meteorological Agency (TMA), and involved with the IPCC. He will be coordinating our new data- and information-sharing collaboration with TMA, hopefully as part of WMO's Global Cryosphere Watch.

As previous entries have mentioned, the past year has been drier than normal at the summit. Decreased albedo has resulted in considerable ablation of both vertical and horizontal surfaces. Indeed, ice loss at the surface caused an unprecedented number of ablation stakes to melt out, and the tipping of several instrument towers. With essential support from our Summit Expeditions (SENE) crew (photos here and here), the towers were reset after 4 nights camped at the summit (see Fig. 2 & 3) and everyone descended safely.

The Furtwängler Glacier provides one illustration of the speed with which glaciers are shrinking on the mountain (Fig. 6). Since February 2000, when Henry Brecher determined the glacier area from aerial photographs, more than 80 percent of this glacier has disappeared. A brief historical perspective on this glacier is available here. The linear rate of area decrease suggests that there will be nothing left of the Furtwängler by 2025.

Many thanks to longtime collaborator Thomas Mölg for helping to support this fieldwork!

Figure 1.  Spencer Hardy and Dr. Ladislaus Chang'a at Barafu Camp (4,670 m), our fifth night of the ascent.

Figure 2.  Looking west over the Northern Icefield. Visible instrumentation includes (left to right) a timelapse camera, high-accuracy temperature and radiation measurement (Climate Reference Network compatible), and the original AWS. Several ablation stakes are faintly visible in the area around the instruments. See next image for detail.

Figure 3.  Northern Icefield instrumentation site at ~noon, looking toward Uhuru Peak on left skyline (2 km distant). This cloud pattern represents typical diurnal development, with convection to the south and west, and rising up the Western Breach.

Figure 4.  Detail of Northern Icefield surface near the AWS, with small nieves penitentes formed since the 2017 long rain season. About 35 cm of the ablation stake is exposed. Between the penitentes is new snow from the previous evening. Also note the area of dirty ice to the right of the stake; the character of all glacier surfaces on Kilimanjaro is spatially heterogeneous and varies tremendously from year to year.

Figure 5.  Rapidly shrinking, east-end remnants of the Northern Icefield, likely once part of an ice body shown in image #95, here.

Figure 6.  The view north from near Uhuru Peak. Northern Icefield in the background, still 40+ meters thick, and the Furtwängler Glacier (foreground); Reusch Crater sloping up to the right. The Furtwängler ice area is 32 percent less than it was just two years ago (Sep. 2015). See image #115 here for the same view in 2013.

Figure 7.  The remaining ice of the former Eastern Icefield, ~1.5 km distant to the northeast.

Figure 8.  Upper Deckens Glacier near Uhuru Peak, one remnant of the former Southern Icefield. Compare with image #33 here from 2009, when the Decken and Kersten Glaciers were still connected. The upper sections of these dirty south-side glaciers provide dramatic evidence for the processes of both sublimation and melt.

Figure 9.  The upper Rebmann Glacier, not far from Stella Point. The recent break-up here has been rapid, associated (in part) with marginal lake formation and drainage; note several areas of buried ice. On the right-hand side of the image, note how the ice stratigraphy more-or-less parallels the slope, yet the ablation surface is nearly horizontal. Selecting sites to obtain ice samples for age dating of these glaciers, or for ice core drilling, is not a trivial issue.

Figure 10.  Looking east from camp, just after sunset. One of the views which keeps us going back!