Lava cascading down the wall of Makaopuhi Crater was hand-colored in this black-and-white photo taken in May 1922. (Historic photo courtesy of Roger and Barbara Myers)

(Volcano Watch is a weekly article written by scientists at the U.S. Geological Survey’s Hawaiian Volcano Observatory.)

In May 1922, the scene at Halemaumau Crater in Kilauea Caldera was spectacular. The level of lava within the crater had been rising since November 1921.

Six months later, upwelling lava and fountains from up to 20 different centers in Halemaumau created lava lakes that reached within 15 m (50 ft) of the crater rim.

On May 13, however, the lava level began to drop rapidly and steadily, and, by May 25, it had dropped almost 150 m (490 ft). Thomas Jaggar, founder and director of the Hawaiian Volcano Observatory at the time, described the fall as “majestic,” as the surface of the lakes, with surmounting crags, subsided as a whole. Swarms of earthquakes — some strong enough to be felt as far away as Hilo — shook the volcano. But that was just the beginning.

On May 26-28, giant collapses of the crater sent ash-laden clouds hundreds of feet high and produced tremendous roaring sounds that were heard several miles away. By the time the collapses ended, Halemaumau had widened by several hundred feet, and the lava column within the crater had dropped at least 300 m (1,000 ft).

Knowing from past experience that lava draining from Halemaumau often fed an eruption or intrusion elsewhere on Kilauea, observatory staff went looking for signs of new eruptive activity. On the night of May 28, a bright glow appeared at Makaopuhi Crater on the volcano’s east rift zone, and an expedition was immediately organized.

Reaching the crater at 3:30 a.m. May 29, they found a long line of vents below the north rim of the crater. Some vents were spattering, while others produced streams of lava that cascaded down the talus slope to form a lava pond at the bottom of Makaopuhi Crater.

Around 9 a.m., observers noticed smoke to the east of Makaopuhi. After trekking through the rainforest for 5 to 7 hours, they discovered “noisy thunderous spouting lava” at the northeast edge of Napau Crater, mostly from the wall of the crater, but also from the plateau above the rim.

Meanwhile, activity at Makaopuhi waned and appeared to have ceased by the end of the day. The Napau phase of the eruption also lasted only about a day. A small amount of lava from this eruption was later found in a crater northeast of Napau.

The impressive summit lava level drop during the May 1922 eruption was not unusual for that time period. Similar drops in level, also corresponding to rift zone eruptions and intrusions, occurred in 1919, 1923 and 1924.

While not as visually spectacular as the massive draining events of the 1920s, the drop in summit lava level during the recent Kamoamoa eruption was just as rapid. We estimate that the level dropped at least 130 m (425 ft), most of it on the first day of the 5-day-long eruption.

Because we now have continuously recording instruments on Kilauea’s east rift zone, as well as at its summit, we know that the summit lava level drop was not the first signal of the impending Kamoamoa eruption.

About 15 minutes before the summit began subsiding, Puu Oo, on Kilauea’s east rift zone, started to subside as magma was drawn toward the eruption site from storage areas in the rift. Thus, we know that the eruption was not driven by overpressure in the summit magma system.

The 2011 Kamoamoa eruption was monitored by a diverse suite of remote instruments that was unavailable to Jaggar when he observed the 1922 eruption. One can only imagine what he would think of the data streams that are collected today and help us understand and ultimately forecast eruptions.