Ana səhifə

Need to know when took text from another source


Yüklə 303.21 Kb.
səhifə1/5
tarix14.06.2016
ölçüsü303.21 Kb.
  1   2   3   4   5

  • NEED TO KNOW WHEN TOOK TEXT FROM ANOTHER SOURCE

  • Need graphic sources

  • Repetitive parts

  • Forecast process – apply it since shows info presented earlier or adjust in some other way?

  • Check for rainfall “performance” – need to be clear what performance is good/bad for

Ignatius:



  • I haven’t done the animation but all the graphics are labeled and well sequenced for animation. To my understanding, animations will be done at another stage but if need be, I can do them.

  • I have now resized ALL the graphics. None exceeds 600x600.

  • Following our recent correspondence with Marianne, I will also avail all the original graphics at a later date. All are available but I have to name (name according to chapter page) them (package) in an easier way to be traced – possibly following the script section numbering (where the graphics belong).

  • None of the animation graphics exceeds 450x450. Most are 400x400.


Early Drought Detection in the Greater Horn of Africa Using Satellite-Derived Data [OK??]
Module Structure [MARIANNE: REVISE]

Section 1: Introduction

About the module

Objectives, Audience and Intended Use

Credits
Section 2: Drought in the GHA



  • 3.0 GHA Countries

  • 3.1 Period and purpose of the study

  • 3.2 Type of the data used

  • 3.3 Chronology of drought in Kenya

  • 3.4 Drought Impacts

  • 3.5 Definition and types of drought


Section 3: Case Study, Data Analysis

  • 4.1 Region of study – the GHA Countries and 2009 NDVI

  • 4.2 GHA-NDVI Anomalies

  • 4.3 GHA Rainfall Estimates (RFE)

  • 4.4 Kenya rainfall performance for 2006 and 2009 – Time series

  • 4.5 Kenya NDVI Time series

  • 4.6 Comparison of 2006 and 2009 NDVI time series for Kenya

  • 4.7 Comparison of 2006 and 2009 NDVI for both Kenya and GHA

  • 4.8 Use of MSG Natural Colour RGB to monitor vegetation

  • 4.9 GHA MSG Natural Colour RGB – 2006 vegetation monitoring focusing on Kenya

  • 5.0 GHA MSG Natural Colour RGB – 2009 vegetation monitoring focusing on Kenya

  • 5.1 SPOT-Global view of very dry Kenya in 2009

  • 5.2 Deteriorated 2009 drought with the failure of the March-April-May Seasonal Rainfall

  • 5.3 MSG Day microphysics channels explaining the severity of the 2009 drought

  • 5.4 MSG 321 Global view of drought stricken GHA in 2009

  • 5.5 Comparing normal NDVI and NDVI Anomalies during the 2009 drought peak period over GHA

  • 5.6 SPOT - Vegetation depression over GHA with ITCZ shift between March and Oct 2009

  • 5.7 Onset of the Oct, Nov, Dec (OND) 2009 seasonal rainfall (Short rains)

  • 5.8 Increased OND seasonal rainfall

  • 5.9 The role of Sea Surface Temperature (SST) in the GHA rainfall performance

  • 6.0 Behavior of Sea Surface Temperatures (SSTs) during a La Nina year

  • 6.1 Behavior of Sea Surface Temperatures (SSTs) during an El Nino year

  • 6.2 The Indian Ocean Dipole

  • 6.3 SSTs and the 2009 Drought over GHA

  • 6.4 SSTs and the 2006 Rains over GHA


5 Drought detection using seasonal forecasts and NDVI as basic data

  • 5.1 Examine the current NDVI and NDVI Anomalies situation

  • 5.2 Look at the Previous Seasonal rainfall Performance

  • 5.3 Look at the rainfall forecast for the next season


6. Effects of the Long drought period

7. Conclusion
SECTION 1: INTRODUCTION

Page 1: About the Module
This module focuses on the use of satellite data, seasonal weather forecasts, numerical weather prediction forecasts, sea surface temperature data, and rainfall data to forecast drought in the Great Horn of Africa (GHA).
[REARRANGE] The first part describes the region, provides an overview of the case study, and describes the types and impacts of drought.
The second section uses data to analyze xxx. [orig: Part (ii) gives the data analyses]
The final part presents a process for detecting drought.
Page 2: Drought in Africa

Drought is one of the most frequent climate-related disasters. It occurs across large portions of Africa, often with devastating consequences for the food security of agricultural households, water supply, crop production, and rearing of livestock. Droughts can lead to famine, malnutrition, epidemics, and the displacement of large populations.
Starvation [mw, get graphics. Loss of Livestock: Source KMD


Drought (and flooding) can significantly impede or erode a country’s economic growth and development.
Drought is affected by a changing climate, and projections indicate that weather extremes, such as drought, will become more frequent.
Drought monitoring and forecasting in Africa is limited by the scarcity of reliable ground-based observational data. There are large spatial gaps between operational weather stations in most African countries, and individual stations often provide discontinuous data.
Temp-upper air observation Stations (Temp Obs.ppt” – single slide in mixed graphics),

Get chart showing synoptic stations from henk!!!!!!!!!!!!!!!!!!!!!!!!! Done


For these reasons, rainfall estimates from atmospheric circulation models and satellite observations are being used as well.
(link to “Map.avi” to see clip). Location: Mixed graphics.
In this case study, we will mainly use satellite observations and NWP model data to bridge this gap (I don’t like this expression) in drought forecasting.
(link to the ppt “Spatial sat data.ppt” – 4 slides)
The goal of the module is to provide students and forecasters with an easy additional method for forecasting drought in the GHA countries, one that does not depend only on scarce data from rain gauge networks. We will use data from these main sources:

  • The Advanced Very High Resolution Radiometer (AVHRR) instrument on NOAA and METOP satellites

  • SPOT satellites

  • The SEVIRI instrument on MSG satellites

  • Model forecasts from the European Centre for Medium-Range Weather Forecast (ECMWF)

  • Rainfall estimates (RFE) from the Climate Prediction Centre (CPC) of the National Oceanic and Atmospheric Administration (NOAA)


The module is intended for use by colleges and universities as well as operational weather services in East Africa, which are frequently affected by extreme rainfall deficits that result in serious droughts.
By the end of the module, students and forecasters should have a better understanding of drought and be better equipped for its early detection.
Page 3: Credits - see bottom of script (only limited information there)
Page 4: About the Case Study: Area of Study and Period of Coverage

The rest of this section provides general information about the region of the case study, the period of coverage, the forecast process, and the types of data used.


The Greater Horn of Africa (GHA) countries include Burundi, Djibouti, Eritrea, Ethiopia, Kenya, Rwanda, Sudan, Somalia, Tanzania, and Uganda.
Geographical Location of GHA: GHA countries2.jpg [get new map with south sudan]


WHY NEED 2 YEARS??? We will examine two years that had very different rainfall performance. 2006 was a relatively a wet year, while 2009 was very dry. NEED SAY THE REST HERE? (not needed) In 2009, dry conditions persisted from January to October/November, when the heavy rains started. They pounded the region into 2010.
Page 4: About the Case Study: Forecast Process and Data Used

There are several ways of monitoring rainfall performance based on the evolution of vegetation. For example, remote sensing provides very good spatial and temporal coverage of the development and amount of vegetation. This is strongly related to rainfall and can be used for drought assessment
Data from satellite channels that monitor vegetation are used in vegetation indices (VI). The Normalized Difference Vegetation Index (NDVI) is a widely used indicator for studying the status of the vegetation and estimate crop production. It’s a simple numerical indicator, which indicates the density and health of green vegetation. The index is widely available in [or from?] meteorological services.
In this module, we will examine ways of assessing the possibility of severe drought in advance using satellite derived data, particularly the NDVI. We will monitor the NDVIs and their long-term anomalies in two years, 2006 and 2009. We will relate them to the seasonal rainfall performance (the amounts received and forecast) and see how to anticipate the possibility of potential severe drought in the GHA region.
Due to the lack of observation data in most GHA countries, we will use satellite spatial derived data for the GHA. For ground truth (“ground truth” is not existing, also data from rain gauges have errors), we will use data from Kenya, which is one of the countries most affected by severe droughts. Therefore, it will serve as a good representative for the other GHA countries.
Adverse rain performance in the region (too much and too little) is highly dependent on several factors:

  • Sea surface temperatures (SSTs), which are related to La Nina and El Nino episodes

  • The behavior of the Indian Ocean Dipole (see: http://en.wikipedia.org/wiki/Indian_Ocean_Dipole)


Thus, our process of forecasting drought includes:

  • Analysing SSTs to determine the possibility of La Nina/El Nino or an Indian Ocean Dipole

  • Studying the evolution of NDVI and NDVI anomalies

  • Examining previous seasonal rainfall performance

  • Looking at the expected rainfall performance for the next season


From this, we should be able to anticipate the occurrence and severity of an impending drought.
The following data are used in the case study:

  • Two MSG RGBs: natural color and day microphysics

  • NDVI and long-term NDVI anomalies from SPOT and NOAA [clarify which satellite]

  • Rainfall Estimates (RFE) from NOAA

  • ECMWF model surface winds (10 metre)

  • Sea surface temperatures

  • Nino indices


For ground truth, we will use the following data:

  • Seasonal rainfall forecasts for the long rains (March-April-May or MAM) and short rains (October-November-December or OND)

  • NDVI time-series ???

  • Reported rainfall from across the country

  • Images of the impact of drought

( I don’t like the expression “ground truthing”. Use a better expression such as: Surface rainfall estimates)

We will use two RGBs:

  • The MSG natural colour RGB for the tenth day of each month; this day was chosen since it corresponds to the decadal NDVI data

  • The MSG day microphysics RGB [say more?]


Page ??: The History of Drought in Kenya

Kenya has a long history of drought and flooding, making it ideal for studying both phenomena. Drought records date back to 1883. The extensive agricultural droughts of the 1980s and early 1990s affected many countries and caused some of the most extreme famine in recent history. The International Disaster Database of the Centre for Research on the Epidemiology of Disasters (CRED) reports more than 0.5 million deaths and 253 million people affected by drought from 1981 to 2010 in Africa.



The table summarizes the major droughts that occurred on the continent from 1980 to 2009. Identifying drought-prone areas and estimating the probability of drought are fundamental for implementing programs that increase food security. The information is also important for determining and interpreting the potential effects of climate change in Africa.
need this?? Some of the major droughts in Africa from 1980 to 2010 [mw: redo since copied from elsewhere]

REGION

1980–89

1990–99

2000–2009

Northern Africa




In Morocco, agricultural output recorded losses in 1992, 1995 and 1997 due to drought. In 1997, Algeria's cereal harvest decreased sharply as a result of severe drought.

The most recent drought in Tunisia and Algeria (from 1999-2002) appears to be the worst since the mid-15th century. That's according to researchers who recently analyzed tree-ring records from the region .

West Africa

The Sahel was hit by a severe drought in the early-mid 1980's.The worst drought in the Sahel during this period occurred during the year 1984 affecting most Sahel countries.







Eastern Africa

The lowlands of Ethiopia and the main productive areas of Kenya have been affected by the 1984 drought. In Ethiopia, the 1984 drought caused the deaths of about 1 million people, 1.5 million head of livestock perished, and population of 8.7 million were affected in all. In 1987, more than 5.2 million people in Ethiopia, 1 million in Eritrea and 200, 000 in Somalia were severely affected.




Rainfall records indicate that, in some parts of the

sub-region, the drought in 2000 was worse than

that experienced in 1984.

Southern Africa

In 1982/83 much of Southern Africa was severely affected

Most of the Southern Africa countries were severely affected by the 1991/92 drought, which was the most severe after the 1982/83 drought. The drought of 1991/92 was the severest on record, causing a 54% reduction in cereal harvest and exposing more than 17 million people to risk of starvation.





Page ??: Types of Drought [modify text from original below]

Drought is defined as a persistent and abnormal moisture deficiency that impacts vegetation, animals, and people. It is classified based on its impacts.




Meteorological drought expresses the departure of precipitation from normal over a period of time. Thus, it describes short-term precipitation deficits. This is one of the primary causes of drought.
Hydrological drought usually defines deficiencies in surface and subsurface water supplies. It affects stream flow, groundwater tables, and reservoir levels. It occurs and recovers over much longer time scales and reflects the effects and impacts of droughts.
Agricultural drought reflects root-zone soil moisture deficits and impacts on crop yields. It is usually expressed in terms of needed soil moisture of a particular crop at a particular time.
Socio-economic drought incorporates water supply and demand for health and economic practices.
need? The first three types are environmental indicators, while the last is a water resource indicator. Meteorological drought is the primary cause of drought. It usually leads to agricultural drought due to lack of soil water. If precipitation deficiencies continue, hydrological drought develops. The groundwater is usually the last to be affected and the last to return to normal levels.
SECTION ??? CASE STUDY: DATA ANALYSIS
WHAT TRYING TO ESTABLISH? WHAT’S THE FLOW? DON’T TELL USERS ABOVE WHAT HAPPENED, LET THEM FIGURE IT OUT HERE. (Agree)
Page 1 NDVI for the GHA in 2009

We will begin our case study by looking at NDVI data for 2009 in the GHA and identifying areas with the greatest anomalies. In the plot, green areas are vegetated areas, cream areas are areas with little or no vegetation; and white areas are clouds.


Need this here? The Great Horn of Africa (GHA) Countries. Figure(2)? In “initial graphics” folder

2009NDVI_loop.avi Animate files in 2009 NDVI”.Location folder: “GHA NDVI and NDVI anomalies and RFE.zip”. (First still image to appear on the animations) name the loop (or? (see “initial graphics” folder for originals) Add legend to graphic


why show a long ppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppppThis NDVI animation goes from?? to ??. (Jan 2009 Dek 1 till Dec 2009 Dek 3 in my animation) (we need a scale explaining the colours as presented later in the module)

In which period did vegetation perform best in the GHA region?



    • Jan

    • April/May

    • Nov/Dec **

    • Aug/Sep

Feedback: The vegetation was greenest in November and December.


Page ??: NDVI Anomalies for 2009
NDVI anomalies are based on long-term averages. This plot [ani?] indicates that most of the GHA region had much lower NDVI values than normal in 2009.
Link to and animate: “2009 GHA NDVI anomalies” in “GHA NDVI and NDVI anomalies and RFE.zip in “initial graphics” folder)
(Scale of colours needed for orientation)


Activity: To identify areas with greatest anomalies in the GHA

FEB

From the NDVI anomalies animation for 2009, the earliest detection of drought should have occurred in the third dekad [10-day period] of February. At that time, southeast Ethiopia, southern Sudan, central and southern Kenya, Somalia, and central and southern Tanzania had below normal or depressed NDVI values. The drought??? values increased in March dekad 1 to dekad 3. [right?] This is important because most Kenyan farmers plant in dekad 1 of March.


APR

The second dekad in April is usually the climax of rainfall in GHA region. Not coincidentally, it’s the most important planting time in central and northern Tanzania.


[redo from other version!!!] But the data show that the area experienced maximum depression in terms of vegetation cover, signifying the presence of drought. The situation deteriorated in Kenya, Ethiopia, and Somalia and eastern and northeast Uganda [SAY WHEN???]. (Northeast Uganda seemed to be less affected as the drought started later, 1st dekad of July, and less severe. Ignatius please comment)
MAY/JUN

South Sudan was very green in May but turned to brownish red in the first dekad of June. (I prefer to talk about “drought” rather than in terms of colours)


Note that Ethiopia continued to display depressed NDVI throughout this period.
GET LEGEND FROM FOLDER


Signs of drought ended by the second and third dekads of October from South Sudan to Ethiopia and then Kenya. Somalia, Kenya, and Ehiopia had well established rains by the first dekad of November. Tanzania remained under NDVI depression until the second dekad of November when the wet season began and continued to the end of the year. (Kenya: 1st dekad Nov. Somalia and Ethiopia: 3rd dekad Oct)

THIS SHOULD BE FIRST! In the animation, what do the colors represent? Select the term that completes each sentence. (agree)
Reddish/brown areas have [increased/decreased**] values of NDVI.

Feedback: Reddish/brown areas have depressed NDVI values, reflecting decreased rainfall.

Green areas have [increased**/decreased] values of NDVI.

Feedback: Green areas have increased values of NDVI, reflecting increased rainfall.



NEW PAGE OR SECTION: GHA Rainfall Estimates (RFE)
GHA Rainfall Estimates
The region received little rainfall for most of 2009 except for some significant (how much is significant?) amounts during the rainfall seasons of March, April, May (MAM) and October, November December (OND).
(Link to and animate “2009 RFE”). Location:“GHA NDVI and NDVI anomalies and RFE.zip”
Use tabs. Link to and animate “2006 RFE”). Location:“GHA NDVI and NDVI anomalies and RFE.zip”
NEED THIS??? Is this a useful ?

Let’s compare the 2009 with 2006. During which months did the rainy seasons occur during 2006?? [which is right? Add fdbk] (Feedback needed from Ignatius)




  • January to February

  • March to May

  • June to September

  • October to December

During which months did the rainy seasons occur during 2009? [which is right? Add fdbk] (Feedback needed from Ignatius)



  • January to February

  • March to May

  • June to September

  • October to December

Which statement describes the rainy seasons in 2006 and 2009?



  • Both years performed equally well Both years recorded close to normal sums of rain

  • Both years performed poorly Both years recorded well below the statistical mean

  • 2009 was better than 2006 2009 recorded more rainfall than 2006

  • 2006 was better than 2009 ** 2006 recorded more rain than 2009.

Feedback: The rainy seasons in 2006 had far better performance than those in 2009. Greenness values were higher over a larger part of the region. [OK?] [VERIFY GREENISH COLOR] (I don’t like the expression “better”. Better is subjective. We have to realize that if a farmer experiences severe flooding, his position is not better than in case of severe drought. Use an expression such as: 2006 experienced much more rainfall than in 2009)
In 2009, which months had the lowest rainfall in the GHA?

  • January to February

  • March to May

  • June to September **

  • October to December

Feedback: June to September had the lowest rainfall in the GHA.
  1   2   3   4   5


Verilənlər bazası müəlliflik hüququ ilə müdafiə olunur ©kagiz.org 2016
rəhbərliyinə müraciət