You are sitting on a warm, tropical, beach drinking a margarita. As you watch the day wane away the sun dips lower on the ocean horizon, and the landscape transforms into brilliant oranges and purples. Behind you the palm trees are bathed in orange, and the landscape has taken on incredible colors with accentuated shadows of even the shortest plant or sandcastle. Almost certainly you bring out your cell phone or camera, because, like all photographers, you find the beauty of the Golden Hour to be irresistible, and you know the peak experience will be short lived. Perhaps you even think to yourself that you wish the beauty of that light could last forever. What if it could?
The Golden Hour is also called the “magic hour” and for a landscape photographer there is no better time to be outside. The terms refer to the period of time when the sun is 6 degrees or less from the horizon. In many regions, like the balmy beach scene above, the moment as the sun sweeps through that 6 degree sweet-spot is relatively short. However, in Polar regions like Alaska, the winter sun has such as a low, southern trajectory, that the sunset-like colors almost never fade.
There are a variety of tools, apps, and websites to calculate the solar angle at your location. I used the NOAA ESRL Sun Position Calculator to determine that in Fairbanks the sun dips to the 6 degree mark on October 24th, 2015 and will remain below 6 degrees until February 26th, 2016. To illustrate the effect of the polar magic hour the images below showcase the colors, and shadows achieved by the low-lying sun. For 3 months, the silver lining of our short, winter days is a luxurious landscape lit by an eternal Golden Hour.
Golden Hour Sunset at the University of Alaska, Fairbanks.
The golden hour casts long shadow, even filling in these fox tracks.
The light of the Golden Hour turns the landscapes into shades of pink, red, and orange.
Long shadows casts by the low sun.
The light of the golden hour pouring through a valley at Angel Rocks, Alaska.
The beginning of the Golden Hour reflecting off the trees and a tributary to the Chena River.
Subtle shades of pink during the Golden Hour.
Shades of pink and orange during the eternal Golden Hour of the Arctic.
The low-lying sun peaking through a downed spruce during the Golden Hour
Pink shades and long shadows in this golden hour shot near the top of Angel Rocks, Alaska
Magic lighting and sunset from Angel Rocks, Alaska.
Beautiful light off the peaks and snowdrifts.
I used several key resources for this article. If you are interested in calculating your sun angle check out :
This entry details a portion of my thesis work at the University of Alaska Fairbanks, and is intended to communicate the findings of that work in a four part series. You are reading part four examining the likelihood of competition between wolves and humans. In order to make the article concise, you may review the general background of this work in part one. I have truncated the background and methods of this work and focused on a portion of the results.
In parts two and three of this series I have been examining where humans in the Yukon Flats, Alaska are traveling to harvest moose and where/how wolves are traveling to harvest moose. A key finding of human access was that humans are mostly operating within 1500 meters of navigable water. During our wolf study I found that travel was based around river corridors. Based on this, I will conclude this series of articles by examining the “Beaver Creek” pack which overlapped strongly with navigable water.
I wanted to begin to understand the likelihood of competition around navigable waters for moose between humans and wolves. Remember, moose exist at extremely low densities and humans and wolves depend on them as a food resource. Therefore, I believe understanding competition is particularly important. To understand the likelihood of competition, I applied my model of human access and overlapped it with wolf locations. I found that 75% of wolf use locations fell within the human access model.
My analysis does not contain temporally overlapping data. Wolf habitat selection may differ in September and October when humans are hunting moose. Wolves could also rely on other prey species other than moose during that period. Also, predation in the Yukon Flats extends beyond wolves. Bears take up to 85% of moose calves each spring. As such, my conclusion is just the beginning research for future biologists in the region. A complete analysis would encompass all predation on moose, be spatially and temporally overlapping, and would evaluate how many moose which are predated could be taken by humans. I hope you have enjoyed this four part series! A full copy of the thesis can be obtained by contacting me. Feel free to do so!
On the evening of December 8th this year, a wonderful series of phenomenon occurred. The sun went down, the aurora remained muted, brilliant stars of the Milky Way dappled the darkness, and a new moon sealed the deal for a night of very dark-skies. I left the orange glow of Fairbanks behind and set off on a quest into the inky darkness of interior Alaska to photograph the Milky Way Galaxy.
When photographing the galaxy you are capturing the “galactic plane” which is the stars which spin out from the “galactic center“. Our sun and solar system reside on the edge of the galaxy, and give us the opportunity to look into it. However, depending on the season and the photographer’s location on the planet, the true center of the galaxy may not be available. In Fairbanks the galactic center would be visible in the summer when it is always light. During the winter the galactic plane of the Milky Way is visible, but we do not get an opportunity to see the center because we are blocked from it by the planet.
Fairbanks has not felt wind for over two months and snow which would ordinary not persist with wind clung to the spruces encasing them . I angled my camera at the bases of those trees and slowly moved at up into the sky after each exposure with the goal of creating panoramic ‘stitches’ of the Milky Way. The method compounds the star density of the galaxy, and brings out distant features like a nebula seen in the upper left of several of the images. I hope you take to opportunity to view dark skies when you can!
The feedback on An Early Christmas Part 1 has been really great, thanks! I wanted to share with you how I have embellished on that first concept of shooting Christmas ornaments under the Northern Lights and also get a bit poetic about the aurora. The aurora this week has been remarkable thanks to a coronal hole from the sun allowing high speed solar winds to reach earth.
I walked out on the ski trails behind my house because the broad and brilliant band of aurora overhead indicated to my aurora-sense it was going to be an early showing. I meandered through snow covered trees maintained in their icy encasement by complete lack of wind for nearly two months. The trail was firm, but as I stepped off my body sunk into thigh deep snow which even though it had fallen 6 weeks ago, was still perfect, soft powder thanks to consistently cold temps. In fact, on this night my breath steamed away at -15F, and a few days earlier I woke up to -23. My anticipation grew as the aurora continued to build in strength and at 10:30 PM an auroral bomb exploded in the sky. The metaphor of a bomb is perfect because it was so sudden that I was caught off guard, and was forced to shoot my camera where I stood in an effort to capture adequately the green and pink shrapnel which rippled and writhed in the sky. The explosion caught me in a towering cathedral of spruces which in the images all point to the source of the disturbance. In five minutes the waves of light ended, but it was only the beginning of series of barrages that kept me awake and in awe until 3AM.
I have been building on the initial ornament concept in a few ways. Although it is difficult to hide a camera in front of a mirror, I am placing it in ways that is not obtrusive. From its hiding place I have shot a full 90 minute star-lapse in the bulb! That image, featured below, is the only one not taken on the night I described. I have also shot a full time lapse in the ornaments which turned out quite wonderful! I hope you enjoy the festive twist on the aurora 🙂
This entry details a portion of my thesis work at the University of Alaska Fairbanks, and is intended to communicate the findings of that work in a four-part series. You are reading part three examining wolf movement in the Yukon Flats, Alaska. In order to make the article concise, you may review the general background of this work in part one. I have truncated the background and methods of this work and focused on a portion of the results.
Wolves are highly studied because they are charismatic, exhibit interesting pack behaviors, and are a key predator in the systems where they exist. Their behaviors including movement speed, movement distances, number of prey killed, and travel distances have been well documented in high prey-density systems, but practically no information exists on these behavior in low or very-low density systems. In an attempt to rectify that, a study was initiated in 2008 to understand the kill rate of moose by wolves in the Yukon Flats, Alaska, where moose are held at low densities (<0.20 per square kilometer) by predation. In an interesting twist, that study found wolves are maintaining kill rates (moose per wolf per day) similar to wolves in high prey density systems. Certainly these results counter what I would predict and lead to a natural question – how are wolves accomplishing such high kill rates in low-prey densities? A known mechanism is that wolves in the Yukon Flats keep small pack sizes to cope with low densities of prey; if you have fewer wolves in a pack, more nutrition is available per wolf during each kill. However, if wolves were traveling further or faster in this low prey-density system was unknown. I predicted that wolves in a low prey density system were traveling further, but not faster than wolves in a high prey-density system to maintain these kill rates. I also predicted they were selecting for river corridors when traveling.
To understand wolf movement, I used the same dataset from the 2008 kill rate study. It was composed of Global Positioning System (GPS) collars on six packs. Thanks to diligence in the kill rate study, I knew where kills occurred along each of the paths. For each pack, I characterized if the wolves were traveling, resting, at a kill site, or revisiting a kill site. These behaviors gave me enough information to calculate the rate of speed they were traveling, the distance they were traveling, the number of days traveling to make a kill, and how long they spent at kill sites. I also using a Generalized Linear Mixed Model to understand what landscape features were important for traveling wolves.
I found some interesting results, and put them in context of 16 comparable research papers of movements of wolves in high or medium prey-density systems. I am presenting the most applicable comparisons from the literature review (i.e., systems where moose are prey and studies where GPS collars were used) here. I found search time was slightly longer and search distance was 2.4 times greater in my low prey-density study area. Search time and search length are correlated together given that wolves are (almost) always hunting when moving. Due to that relationship, the search time is expected to go up as search days goes up. I found no evidence that wolves were handling prey longer or traveling faster in the low prey-density system. Those results were not surprising as one researcher found that handling time of moose was not significantly different among packs which varied in size from 2 – 20. Since wolves were not traveling faster in our system, it is probable that regardless of prey density, that on average wolves travel at their maximum comfortable speed that maximizes efficient travel.
I also found that wolves were utilizing river corridors and that they were selecting strongly against brushy habitat. In the Yukon Flats, that means they were selecting against thick stands of alder and willow. This was similar to previous studies where they found that wolves were able to travel 2.8 times faster if they used a river corridor rather than moving through a brushy environment. By using rivers, wolves were traveling faster and are likely taking advantage of increased prey density along river corridors.
The results of this work have some useful applications in helping us broadly understand wolf behavior. First, wolf territories are very large within low prey-density wolf systems. The mechanism that creates these large territories was unknown, but long-distance movements by wolves would create large territories by default. Next, back in the mid 1980s a researcher suggested that 0.20 moose per kilometer squared was the lowest density that wolves could persist at. Within the Yukon Flats, they are already persisting at lower densities than that, and since they are able to extend their travel distances to maintain kill rates it seems a minimum prey-density threshold could be much lower. A final implication of this work is that managers should expect wolf territories to increase in size if prey density decreases. In other systems (for instance deer in the mid-west), wolf territories should inflate in size as they move further in search of prey.
I look forward to presenting part four to you soon, which ties together moose hunting by wolves and humans by starting to understand the likelihood of competition.
This entry details a portion of my thesis work at the University of Alaska Fairbanks, and is intended to communicate the findings of that work in a four part series. You are reading part two. In order to make the article concise, you may review the general background of this work in part one. I have truncated the background and methods of this work and focused on a portion of the results.
How do you get to a resource? Well, the simple answer is you “access” them. Depending on what you are trying to achieve, access may mean walking through the door of your local grocery store, driving onto a frozen lake and drilling a hole to jig up a fish, or driving a boat up a river to harvest a moose. The last example speaks directly to subsistence use patterns of communities in the Yukon Flats, Alaska. The objective of this part (specifically Chapter 1) of my study was quantify rural hunter access in Alaska.
Let’s take a step backward quickly to look at why access matters. Game levels are traditionally managed to create yield for hunters, but it is critical that game populations be accessible to hunters. In the huge area of Alaska, creating high game densities in a remote region may have minimal benefit to hunters. Outside of Alaska, the effect of access on game populations and hunter success is not well understood, but increased access in Ontario may decrease moose, increased access in Idaho may increase elk mortality, and hunters in Minnesota concentrate their efforts within 0.8 km of roads 98% of the time. These studies suggest that access is important, but within the Arctic access has not been quantified despite being important for hunters, particularly those with a subsistence lifestyle.
It is important that game managers understand how many animals are being harvested to aid in setting regulations. In Alaska, this is accomplished by reporting harvest via a “harvest tag”. However, under-reporting of harvest via the harvest tag system is high in the subsistence communities of the Yukon Flats. This is due to a variety reasons centering around culture practices and feasibility of reporting. Within those communities, moose hunters are allowed one bull moose per season, and hunting most often occurs along rivers in September and October.
To understand where moose hunters are harvesting moose, I used an interview dataset collected in 2005 and 2007 by the Council of Athabascan Tribal Governments. The interviews were in conducted in five subsistence communities including Fort Yukon, Beaver Creek, Circle, Arctic Village, and Birch Creek. In the interview process, interviewees recorded harvest locations of moose on a topographic map. Based on that we determined they utilized rivers, a hunting method that is well documented in other research. However, the data allowed me go beyond just determining river use. I wanted to know : how far were users traveling from their community and from the river to harvest moose?
I designed a method to quantify hunter access. I measured the straight-line distance of the harvest points from their community of origin, and the distance from the rivers. The idea behind this is that the hunter moved up river to a certain point, and then moved away from the river a certain distance. I grouped the resulting distances into five groups, and created a buffer around communities and rivers based on those distances. Within the buffers, I developed an “access index” with the goal of understanding the likelihood that a hunter would utilize an area. The access index was calculated as the number of points that fell inside of a buffer divided by the total number of points up to the edge of that buffer. So, based on that the maximum achievable value was 100% and either existed near community, or near the rivers. In effect, 100% means that 100% of the time, hunters were willing to travel that distance to harvest a moose.
The approach that I took was novel, and yielded some useful results. We found that on average hunters were traveling 0.9 ± 0.6 km from rivers and 47km ± 32km from their communities. Harvest was centered around rivers, and was happening most frequently near rivers. Some useful results!
There are a few ways that this model may be applied. First, I applied a region density of 0.0016 bull moose per square kilometer (remember, there are VERY low moose densities) to estimate the number of legal moose that are available to moose hunters. Based on hunter success of 27 – 46%, I estimated that 98 – 176 moose are harvested by hunters annually. Those numbers fell into the reasonable range of reported harvest in the region. Seeing as that’s the case, this method could help managers understand the amount of moose harvested, instead of relying on the extremely (regionally) variable harvest ticket system. Since this model enables an estimate of the number of animals taken around an access corridor, it could be used in other hunting systems where access is important. For instance in Alaska if a new road was created, how many moose would be harvested based on the new access. In Idaho, how many elk would be preserved if a road is closed?
Overall the results of this study have applicability within my study system, other subsistence systems in Alaska, and more broadly to regions where harvest of game is linked to access. It demonstrates a novel method, and the results that can be gained through an interview process. In the next portion of this series, I will be examining wolf movement in this same area, which yielded some great results.
*The entirety of this work is in review with the Journal of Human Dimensions of Wildlife