In arid regions where people are not able to settle and obtain the most suitable means of monitoring ecological element on the land, it is Remote Sensing that provides raster data due to its application of sophisticated satellites. According to Morain (1996, p. 67), the satellites capture and measure the rate of reflection in the electromagnetic spectrum of visible, infrared and microwave lights. Every satellite has a precise ability of monitoring depending on the date they were launched and their purpose. For example, spot satellite managed by the French space agency mostly traps and broadcast data for regional cartography and planning, agriculture, land use, vegetation and capacity of trapping waves of long-length such as infrared and red light waves with a superior resolution as compared to Landsat 1-5 (Wiegl & Giraudon 1993, p. 187).
Meanwhile satellites of Landsat which are manned by National Aeronautics and Space Administration in the United States of America have a superior imageries resolution which make the main sought source of making decision (Rustad 2012, p. 176). Additional Landsat feature is its observation of distance database as from early 1970s (Scanln 2010, p. 223), having Landsat 1 being established in 1972, with two imagers of viewing the Earth. Landsats 2 and 3 were launched in 1975 and 1978 in that order having similar to the Landsat 1 configurations. Landsats 4 and 5 came in 1982 and 1984 respectively having Multispectral Scanner (MSS) and the Thematic Mapper (TM) (Menenti 1986, P. 145). Landsat 6 came in 1993 but failed to revolve in the space and was instantly lost. Landsat 7 which came into space on 15th April, 1999, having the Enhanced Thematic Mapper Plus (ETM+) aboard with a five year lifespan is still in action (Menenti 1986, P. 146). The Landsat Data Continuity Mission was scheduled to be launched before the 11th of February, 2013 and was advertised as the Landsat satellite technology of the future (See figure 1 below).
According to Loveland, (2007, p. 76), the LDCM will have an exceptional aspect of obtaining continually valuable imagery and data to be used in government, science, agriculture, education and business. This will be enabled by its ability to attain repetitive high resolution acquisition of multispectral data of the surface of the earth on a global foundation. The information acquired from the spacecraft of the Landsat includes the longest data of surfaces of the earth as observed from space. Loveland, (2007, p. 106) notes that this data is not parallel in quality, value, coverage and detail to any other sources of satellite. Additionally, Landsat provides image data for Normalized Density Vegetation Index (NDVI) that is an efficient method in determining change in vegetation over period together with classification of vegetation. This method makes use of the visible and near-infrared electromagnetic bands of spectrum customized to determine the measurements of remote sensing and scrutinize the under study phenomenon has green live vegetation (Holme, Burside, and Mitchell, 1987, P. 245). NDVI has a broad application in studies of vegetation since it is frequently applied in pasture performance approximation, carrying abilities of land and crop harvest besides other uses. Usually, it is associated with other factors of land such as the activity of photosynthesis of plants, ground cover, leaf area index, surface water and amount of biomass. For the first time, NDVI was used at the Remote Sensing Centre of Texas A&M University by Rouse along with his team (Roderick, Smith and Ludwick 1996, P. 126). NDVI is calculated by use of the below equation;
(IR - RED) / (IR + RED), whereby IR is the infrared and RED represents the red light. Landsat Thematic Mapper (TM) 4 and multi-spectoral scanners (MSS) 6 and 7 have the largest sensitivity strength to sense reflectance of infrared emitted from cell of plant which are changed by accessibility of water (NASA, 2011). Disappearance of vegetation cover is believed to be a result of several factors such as problem of drainage, stress of water, senescence and diseases caused by pathogens besides others. Since it is difficult to determine the actual cause, it is very important to investigate the possible factors that consequence in disappearance of canopy to be able to preserve the surrounding for sustainability of ecosystem.
From the analysis of vegetation vigor using disparity between NDVI 200 and 2001 dieback of plants in Flinders Range Park’s Trezona Range was set up. It is a basic setback since diminishing natural vegetation cover can consequence in exclusion of animal species which depends on the vegetation (Liu and Kogan, P. 234). Therefore, this research seeks to determine the association between vegetation disparities; this is established from possible causes of dieback of vegetation together with NDVI raster data.
Area of Study
Flinders Ranges National Park’s Trezona has two areas of study situated 450 km north of Adelaide, with both having an area of 1.2 square kilometers. Monthly records of climate at HAWKER, the nearby station, show that yearly evaporation in roughly 2500 mm and approximate annual rainfall is 300 mm (Liu and Kogan 1996, p. 232). The region has clear wet winters and dry summers and, thus, termed as a semi-desert area. A large part of the region is covered with thick deposits of sedimentary rocks which stretch to Adelaide Fold Belt (Gostin 2011, p. 128). Flinders Range has about 123 exotic plants species making a melting pot for studying conservation of environment.(See figure 2 below)
Methods and Materials
An Arc Map and Arc Scene were used to examine satellite images (NDVI2000 and NDVI2001), with a 25 meters resolution to determine transformation of vegetation. The two photos were taken in July 2000 and 2001, covering areas of 9.6 km by 9.8 km. The disparity in the plants in the study area from 2000 to 2001 was calculated by use of raster machines and mask function. To reach at the association between transformation of vegetation and possible causes, zonal statistics tool and recluse function was used. The possible causal aspect of slope direction, condition of slope and distance from streams were considered too. In addition to that, climatic data was derived from Blinman station which was also applied to review disappearance of vegetation canopy in the area. The zonal function facilitates data transmission from one layer to another. It requires two layers: zone layer and value layer. Single layer data is calculated centered on zone layer. Zone layer can highlight locations, shapes and values. Distance from streams, reclassified slope and elevations were used as key criteria for obtaining raster data, while transformation of plants in the study region was used as the key criteria for raster value. Because of this process, every data such as means, minimum, standard deviation and maximum were calculated based on the raster input classification. See figure 3 below.
Figure 4 below highlights change in vegetation between 2000 and 2001. Areas of study are indicated by black lines. Healthy plants are symbolized by green cells, whereas red cell represents fall in the vegetation cover. In addition, with regard to transformation of vegetation in the area of study, statistical data were contrasted with that of captured area (see table 1 below) clearly shows that vegetation loss is occurring in the study areas. The table depicts vegetation loss by use of negative aggregate value of -0.029, whereas 0.005 value of standard deviation was as well recorded.
Effects of Slope to Change in Vegetation
Figure 6 below denotes the effects of slope to the change in vegetation having two classifications of slope; below 10% and above 10% vegetation change on land (within the area of study) whose steepness is below 10% is 0.024, while transformation on land with steepness of above 10% is -0.044. Figure 7 below also comprises two classifications; below 10% and above 10 % vegetation change on land (outside the area of study) whose steepness is below 10% is 0.008 whereas transformation on land with steepness of above 10% is 0.003. Generally, steep slopes are inappropriate for plant growth regarding factors of hydrology; the steeper the slope the faster the ground water flows with a higher potential of degrading the land; this proportionate affects vegetation degradation. This conclusion is upheld by results of climate studies on seasonal variation. Variation through seasons also results in loss of plants and affects the growth of vegetation; this resembles the scenario of the area of study because of the presence of two different seasons (hot dry summer and cold wet winter). Ecosystems of plants in the study areas comprise ephemeral grasses, drought resistant shrubs as well as schlerophyllous trees. Each NDVI photo was taken during July marks the beginning of wet season; thus, transformation of the beginning of wet season is probable to affect plants cover because grass grows faster as compared to trees. Aggregate monthly records of rain as showed in Table 3 illustrate a notable difference of rainfall since the month of April up to July 2000 as well as 2001 respectively. The difference in amount of rainfall was comparatively higher in 2001(4.2 mm to 77.6 mm), whereas the rainfall amount in 2000 was equally distributed (Liu and Koran 1996, p. 254). This implies that tremendous patterns of weather consequence to negative impacts on growth of vegetation specifically areas with steep land due to soil degradation and landslides whereas on the flat slops grass grew due to plentiful water.
Impacts of Distance Creeks on Change of Vegetation
Regarding association between change in vegetation and distance from streams, the average is conflicting to the outcome from the conclusion resulting from the slope analysis. The vegetation change of the region positioned at 100m from the stream up to regions located 400m from the stream are shown in Figure 7. Vegetation change for a location inside the radius of 100m from creeks was 0.005 while 200m from streams was -0.021. Negative 0.027 and negative 0.036 were recorded for distances 300m and 400m respectively from the creeks. The propensity seemed to be similar though the study area spurned though the area captured. Table 4 shows that the aggregate vegetation change inside the radius of 100m from streams varies from 0.014 to 0.025 in the areas positioned inside the radius of 400m from creeks, although, the disparity is not very definite. Incidentally, when aggregate annual rainfall data for July 1999 and June 2000 to June 2000 and july2001 was contrasted, it resulted to NDVI2001 being 376.7 mm/annum whereas NDVI2000 being 284mm/annum implying that vegetation loss in the area of study was not due to stress of water. (See Table 3)
GSI through the use of Remote Sensing Data has an important function in the stipulation of spatial information using the modest resolution in regions which have poor accessibility in a lesser time and comparatively lesser cost as compared to data from analysis of field work. GSI can as well offer spatial distribution of soil type that is among the main indicators for change in vegetation. Both aerial satellite images are helpful in determining disappearance of vegetation. Thus, color infrared aerial snaps (CIR) are helpful in the forest research dieback. This has confirmed to be timely, inexpensive and accurate technique used by Croatia (UN-ECE EC 2013, P. 323). Besides offering spatial data at cheap cost, CIR points the casual effects of meticulous variables by integrating the layers of data and examines them. Moreover, it can be used to forecast future trends of phenomena harmonizing efficient fieldwork and suitable response to it. From the association data, the major source of dieback of forest could be extreme rainfall in June 2001, that resulted to degradation of soil and vegetation in steep slopes areas, whereas relatively flat regions experienced an improvement of growth of vegetation due to minimal impacts of extreme rains.
However, this outcome depicts vulnerabilities of analysis NDVI, for instance, grass usually grows faster compared to trees while NDVI DATA cannot identify the types of trees. Dieback of forest mostly influences trees, but NDVI analysis misses out this because outcomes do not properly illustrate distribution of dieback. To mitigate the NDVI results setbacks, hyper-spectral imaging can be applied to classify types of vegetation. This technique offers data which has spectral arrays much close as compared to that of NDVI, which is capable to portray particular types of vegetation from values made in a laboratory (Cloutis 1996, p. 221). In flat regions, vegetation is mainly grass and the possible cause of loss of vegetation is attributed to intensive weather conditions. Grouping of vegetation can facilitate differentiating various causes of vegetation loss such as dieback, burning insect-caused damage, senescence and types of soil. This method has, thus, been applied broadly in researching change of vegetation. Bravo (2004), detected yellow rust on wheat with a minute error margin (5 - 6%) by use of combination of hyper-spectral imaging and analysis fluorescence image. He further found out that hyper-spectral imaging is most appropriate for sensing rust of leaf.
Besides the abovementioned techniques,fieldworks help to find out causes of vegetation loss, although they are expensive cost and time consuming. For instance, when studying quality of underground and surface water and depth of water table, the major important variables to consider dieback are samples of trees to obtain detailed information through conducting reflectance spectral in laboratory.
GSI is among powerful tools which can be efficiently used to sense vegetation change over a specified period at low cost and minimum time. Utilizing of spatal 3D and 2D mapas together with analysis of raster data, allows to arrive at a conclusion that the causes for dieback of plants are due to intensive weather condition in June 2001. Lastly, there is a neccesity for further research to obtain detailed information on water table and water quality using hyper-spectral imaging and fiedwork methods.