Successful FEWS NET UCSB/Field Scientist Collaborations

Here we briefly provide two examples of climate-related early warning activities in 2002/2003. Both cases involved collaborations between USGS scientists in the U.S. (at EROS Data Center and UCSB), USGS/FEWS NET representatives in Africa, and collaborating institutions in Eastern and Southern Africa.

Somali Flood Mitigation (Fall 2002)

This story begins in early October 20021 with the weekly African hazards teleconference hosted by the CPC and comments by Wassila Thiao and Jim Miller noting the flood risks implicit in a very warm Indian ocean. Exploratory data analysis with October-November rainfall for the region confirmed this result, and suggested that conditions in the upper atmosphere were also conducive to supporting heavy convection. Standardized precipitation index images based on MRF forecasts provided by Greg Husak at UCSB helped confirm the risk potential. The analysis of the Madden-Julian oscillation and COLA precipitation forecasts provided by NOAA/CPC all suggested that heavy rains might be a possible risk in Somalia. This information was shared with FEWS NET hydrology experts (especially Guleid Artan and Hussein Gadain). Because of extensive networking activities by Hussein Gadain and Sidow Addou (They attended at least 18 meetings on the subject between April 2002 and Jan 2003), FEWS NET had established excellent contacts with the flood monitoring community in Somalia. Hussein is a member of the Flood Working Group (FWG) of the Somalia Aid Coordination Body (SACB). Sidow Addou, the FEWS NET Somalia representative played a very critical role in monitoring conditions and contacting preparedness experts in Somalia. FEWS NET Somalia issued a weather advisory (Figure 1) showing forecast precipitation maps from the CPC and Basin Excess Rainfall Maps from the USGS (Figure 2). This information was picked up by humanitarian agencies, such as UNICEF, and motivated flood preparedness activities. These activities involved the identification of vulnerable populations, identification of safe havens for IDPs, the assessment of available emergency supplies, and the development of a Somali monitoring system in the Shabelle and Jubba basins2. These preparatory activities were well founded. The following news article3 from November 11th described the situation.

Large areas of the Juba Valley in southern Somalia have been flooded, with thousand of hectares of farmland inundated by flood waters, according to sources in the affected region.
Businessman Abdinasir Abdullahi Haji from the town of Afmadow, 620 km southwest of Mogadishu, told IRIN the floods had destroyed farms and properties in the Lower Juba and Middle Juba regions. "In October we had two weeks of non-stop rain," he added.
He said the Deyr rains (September -December) started early this year, cutting off all roads and severing business activity between the town and Kismayo and Mogadishu. "Prices of essential goods have quadrupled," he said.
However, the UN's Food and Agriculture Organisation (FAO) said the situation did not appear grave…

The information provided by FEWS NET partners working together and distributed effectively through local institutions helped mitigate the impacts of this flood.

SADC Drought Forecasts

This story begins in September of 2002. A special report by FEWS NET, in collaboration with the SADC Regional Remote Sensing Unit, used a newly available 40 year climatology of Water Requirement Satisfaction Index imagery to map the potential dangers associated with the development of a moderate-to-strong El Niño event4 . Tamuka Magadazire contributed substantially to this work, as did members at EDC (Gabriel Senay, Jim Rowland and Jim Verdin). The study highlighted potential problems in the ‘Drought Alley’ in Eastern Botswana, Southern Zimbabwe and Southern Mozambique – a region previously identified as a high risk area at a Drought Hazard Mapping workshop (see Figure 3). Since Zimbabwe faced dramatic food insecurity, it was prudent to highlight the possibility of repeated drought – an uncommon event for Southern Africa. These sentiments were also shared by collaborators at the RRSU (Elijah Mukhala and Kennedy Masamvu), and entered cautiously into several ministerial briefs.

In mid-October the El Niño intensified and rainfall was very heavy. A special report (primarily authored by Greg Husak) showed a very strong negative link between early (October) heavy rains and drought in January-February-March during El Niño years . The relatively light impacts of the recent 1997/1998 El Niño year, in combination with the heavy early rains, over-mollified fears of ENSO-induced drought impact on food security. The FEWS NET team put together a carefully cross-validated statistical model which was used to forecast end-of-season maize WRSI values for eleven regions (Figure 4) . These forecasts helped anticipate the observed drought conditions, correctly identifying most of the at-risk regions (Northeastern Republic of Southern Africa, Southern Mozambique, Southern Botswana). This report was followed up by two innovative reports from Southern Africa that estimated the likelihood of maize crops receiving sufficient rains in February-March-April (Figure 4), using forecast precipitation maps and the Forecast Interpretation Tool (FIT) developed at UCSB (Figure 5). These maps were incorporated into reports issued to key government figures, demonstrating that innovation can span the gap between R & D and decision makers.

Taken together, the FEWS NET reports provided a fairly complete picture of the drought risk, beginning with climatological risks (the DHI), proceeding to ‘passive alert’ stage based on historical El Niño teleconnections4, proceeding to ‘active alert’ stage based on observed climate conditions early in the season5,6, proceeding to ‘early drought assessment’ activities at mid-season7.

While these stories represent only a portion of the work done by FEWS NET, they do provide good examples of effective early warning of hydrologic risks distributed through networked institutions. When all the pieces are put together we can anticipate the short-term vagaries of climate, issue early warning, and motivate effective response.