If It's Not One Thing, It's Another
Weather and Climate for the Roanoke Colony

Dennis B. Blanton

Any colonizing effort faces risk, and some elements of it are better calculated than others. Weather surely ranks among the most difficult factors to predict, and the late-sixteenth-century English experience in the New World was hardly sufficient to anticipate well the nature of climatic norms and extremes on what is today’s Outer Banks. Certainly the thin barrier strand presently bears a notorious reputation for weather extremes, especially in the form of warm-season hurricanes and cold-season northeasters. Only regular firsthand observation over at least many decades permitted recognition of such dangerous extremes typical of the Roanoke area. And even under the best of circumstances unanticipated challenges can arise.

    This paper offers a look at some of the climatic factors that affected early English colonization, especially during the late sixteenth century’s several attempts to secure a foothold on present-day coastal North Carolina. Some pieces of this picture are relatively familiar and, as noted, define the reputation of the Outer Banks as the “Graveyard of the Atlantic.” Other pieces, such as the effect of severe drought, have emerged recently. Together they remind us that the natural world often matters a great deal in human affairs.

A Startling Reminder

    Roanoke-related climate held little interest for me before about two years ago, when I, along with colleagues from the University of Arkansas’s Tree Ring Laboratory, found that subject suddenly thrust into our lives. We serendipitiously identified evidence of a climatic extreme—a remarkably severe drought—precisely at the time the final Roanoke party arrived in 1587. Our original purpose had been to explore conditions associated with the ultimately successful English colony at Jamestown in 1607. David Sthale and his colleagues, specialists in dendroclimatology, use annual tree-ring patterns of ancient, living bald cypresses to reconstruct past conditions. I had enlisted their help for the Jamestown study and had even formed a hunch that drought might have contributed to that colony’s struggles. Indeed, the Arkansas analysis established that a significant drought did coincide with the period 1606-1612 and must have played a role in the now legendary hardships there.

    I recall well our early, excited conversations about the Jamestown finding. Almost as an aside, Stahle called my attention to a major “blip” in the eight-hundred-year-long tree-ring record from southeastern Virginia and northeastern North Carolina. He reported that, statistically speaking, the absolute worst drought of those eight centuries occurred between 1587 and 1589. We were both so taken by the revelations of the Jamestown record that several days passed before we connected this most severe drought with Roanoke! Suffice it to say that it was a startling realization that drought conditions had handicapped not just one but two of the earliest English colonial settlements. It is a circumstance whose implications now warrant careful consideration.

Understanding Atlantic-Area Climate at the End of the Sixteenth Century

    Climate, defined as the long-term patterns of atmospheric conditions of a given area, is a fact of human existence everywhere but generally gets taken for granted—it is among the conditions to which a culture adapts and its members become accustomed. This is not to say that the effects of climate are trivial, since it is essential that human societies adapt over the long run in order to sustain themselves within their environment. Indeed, the adaptive patterns of a society can virtually define its culture, or significant elements of it at least.

    English-speaking Europeans were only vaguely familiar with climatic conditions in “Virginia” and could not accurately anticipate how the differences in climate could affect their activities. Even after they arrived, their first impressions of the local environment were not always accurate for a variety of reasons. I summarize below the conditions familiar to the men and women who left England in order to demonstrate how the new Virginia climate challenged them.

    Hindsight helps us all, and paleoclimatologists make it their business to look back, determine the nature of past climatic conditions, and apply the knowledge to the present and future. Climate historians have recognized the onset of a cooling period in Europe that extended variably from the twelfth through the mid-nineteenth centuries. This period probably did not, however, feature uniformly lower temperatures but, instead, more frequent intervals of extended cold than at present. Moreover, the episodes of cooling also did not necessarily affect the whole of Europe each time; often they were regional. Detailed reconstruction for England in the late sixteenth century were leaving a generally cooler climate and that the periods 1560-1570 and 1590-1610 were among the coolest. The relatively high-frequency periodicity in the records indicates that during this so-called Little Ice Age the climate was more subject to extremes, and Swiss data suggests that the period around 1600 was the height of that trend.

    Given the paucity of records, conditions across the Atlantic in eastern North America during the 1580s are even more difficult to measure. The dendroclimatological data reported here is one segment of a long record that can provide a broader view of regional conditions not available before, but, as indicated, they are most pertinent to precipitation reconstruction. Cumulative tree-ring records from all of North America indicate that a prolonged drought plagued the continent in the late sixteenth century. That dry period applies to the Carolinas and Virginia in the 1580s. Ice-core records for northern Canada indicate that conditions there were markedly colder from 1550 to 1620. Well to the south, South American research records a moderate to strong El Nino event between 1589 and 1591. The potential effect of this pattern in present-day North Carolina is somewhat milder and wetter winters and a lower frequency of hurricanes, although the effects of El Nino on the Middle Atlantic region are not yet well understood. As Stahle and others note, both hurricane frequency and the likelihood of drought rise during La Nina years, and such conditions may have prevailed at Roanoke when the English push was under way. Until more thorough reconstructions are available for the eastern United States, these far-flung hemisphere patterns only suggest general conditions.

Specific Factors in the Roanoke Story

Drought at Roanoke

    Bald cypress (Taxodium distichum) tree rings provide an extraordinary record of past climate for the area of the Roanoke colony. Millennium-old trees not only extend the chronicle well back in time but also translate into an independent, quantifiable gauge of the characteristics of historical climate. It happens that these trees are very sensitive to trends in growing-season precipitation, which each year extends between about April and July, especially when they grow in sedimentary as opposed to peaty substrates. Cores from several living trees are analyzed, and ring measurements are combined into a single, calibrated sequence for the region. By standardizing the ring record, a series of indexes can be applied to the sequence in order to measure relative severity of droughts and other trends.

Fig. 16. Graph showing growing season soil moisture in the Roanoke area from 1580 to 1600, derived from tree-ring records and measured by the Palmer Hydrological Drought Index.

    The tree-ring record for southeastern Virginia and northeastern North Carolina generated by David Stahle and his team at the University of Arkansas was full of surprises. Most notable is the severe downturn in rainfall during the years 1587-1589. Measured statistically, it amounts to the deepest low in the entire eight-hundred-year story of regional precipitation trends (Figure 16). Although it is tricky and sometimes misleading to offer modern analogies for such events, it is reasonably safe to say that this Roanoke-period deficit exceeds anything experienced in the region in the late twentieth century. A drought of this magnitude, though short-lived would affect human welfare in these pre-modern societies.

    Almost always, severe drought most directly affects water supply and food supply. Precious little corroboration evidence can be gleaned from the scant records left from the Roanoke period. But there are occasional suggestions of drought-induced hardship. The clearest comes when John White describes the initial reunion of the English and Indians during July 1587. White relates that the natives immediately expressed concern about their lack of corn: “Some of them [the Indians] came to us, embracing and entertaining us friendly, desiring us not to gather or spoil any of their corn, because they had but little.” Establishing the certain and precise meaning of these recorded statements is among the more difficult challenges historians face, but one would be remiss to dismiss the local Indians’ expressed concern as entirely deceitful or calculating, especially in the light of the new and independent evidence from bald cypress trees.

    The water question is more difficult to pin down. Fresh water is perhaps the human requirement least to be taken for granted on the barrier islands. The sandy soils of these thin coastal strands not only drain quickly but have limited reservoirs of potable groundwater accessible to shallow wells. Drinkable surface water was likewise scarce, probably even in the best of times, as John White’s 1590 account of an island exploration following the drought period attests: “[N]or had we found any fresh water in all this way to drink….The sailors in our absence had brought their cask ashore for fresh water, so we deferred our going to Roanoke until the next morning, and caused some of them to dig in the sandy hills for fresh water—which we found very sufficient.” Suffice it to say that a drought very likely all but erases surface water sources and reduced the volume and quality of shallow wells in the sandy island soils.

    The fate of the “Lost Colony” population remains one of the great mysteries of the nation’s history. The introduction of drought as a potential factor leading to the demise is warranted now, but it is not sufficient to account for the disappearance singlehandedly. The lack of rainfall probably introduced more stress into a situation already stressful for both the Indians and the English. The apparent food shortfall cited earlier could well translate into competition for a suddenly limited food staple for both groups and thus, engender conflict as the two sides struggled to control the supply. Water problems would only have exacerbated the quandary. Under another scenario, the colonists may have opted to “go native” and adopt the Indians’ traditional survival tactics during times of crop failure.

    It is worth noting that tree-ring and other records suggest greater-than-average overall climatic variability during the late sixteenth century, including significant swings in precipitation. One would wonder how, otherwise, each early English colonizing attempt in the Middle Atlantic could have confronted dry conditions. Nor was it only the English who suffered as a result. As noted, compelling evidence suggests that the Spanish settlement at Santa Elena in present-day coastal South Carolina encountered serious drought-induced food shortages.


    The intrepid English struggling to establish a presence at Roanoke perceived impressive storms much more readily than drought, and they left a vivid record of their experiences with them, as would any seafaring people. Historical climatologists refer to hurricanes as extreme events, in this case short-lived but potentially very influential natural phenomena. Even in more recent times, severe storms have sculpted coastal geography on the Outer Banks, creating various landmarks, including Oregon Inlet in 1846.

    At present the defined hurricane season for the section of the Atlantic coast occupied by the Outer Banks extends from June to the end of October. Hurricanes occurred in the sixteenth century as they do today, but their frequency and intensity then are less certain. Some storm researchers have suggested that hurricanes are subject to cyclical patterns over periods spanning several decades. The hurricanes that Gilbert, Raleigh, and Drake encountered off Roanoke were among the first to be described, and since 1583 hurricanes documented as tracking near or across the present-day Outer Banks number in the hundreds. During the 1583-1590 English colonization at Roanoke, at least four probable hurricanes were described off the North Carolina coast. In the age of exploration, the preferred sailing season overlapped the hurricane, and sailors and settlers frequently encountered the powerful storms. To travel in warmer weather with favorable winds made the risk worthwhile.

    One such apparent hurricane directly affected the course of events at Roanoke. A storm of June 13-16, 1586, as described by Ralph Lane, forced abandonment of the first settlement. Its four-day ferocity is described dramatically as follows: “[T]he storm drove sundry of the fleet to put to sea…The weather was so boisterous and the pinnaces so often aground, that most of all we had…were cast overboard by the sailors—the greater number of the fleet being much aggrieved with their long and dangerous abode in the miserable road.” This bout with a hurricane facilitated the decision to abandon the initial settlement after about a year rather than stay to await resupply.

    Less pivotal but probably no less troublesome were other great “tempests” that hampered colonial progress. Soon after White established a settlement in the summer of 1587, a storm “at northeast” complicated Sir Francis Drake’s landings for six days in August. The streak of bad luck continued when two more storms interfered with White’s long overdue resupply trip in August 1590. In a 1593 letter to Hakluyt, White’s disappointment and frustration with the last storms are clear: “When we arrived there the season was so unfit and the weather so foul, that we were constrained to forsake that coast, without having seen any of our planters. We lost one of our ship’s boats, seven of our chief men, and three of our anchors and cables. Most of our casks with fresh water were left on shore, not possible to be had aboard.” The rest of the story is all too familiar.

Lessons to Remember

    This examination of climatic conditions at the time of Roanoke colonization cannot solve the “Lost Colony” mystery, nor is it intended to reinterpret the events from a wholly environmentalist standpoint. The aim, rather, has been to introduce these factors in a systematic fashion so that climate can be respected among the many contributing elements of historical analysis. In the present technological age, we are buffered from environmental extremes in ways unimagined before—to the point that it is easy to dismiss climate as a legitimate agent of cultural change, especially during the more recent historical period.

    Another purpose has been to demonstrate the usefulness of an explicitly interdisciplinary approach to the study of history. Little of what is presented here could be possible without the benefit of collaboration with natural scientists concerned with climate change. Introducing these factors, I believe, produces a richer, more accurate interpretation of the riveting events surrounding the Roanoke colonies.

Note: A Table and Notes follow essay in original volume.

Return to Selected Essays

   Copyright © 2007-1999 ~ All Rights Reserved