Remote sensing applications for the assessment of the geomorphic response of fluvial systems to the Holocene Climate Changes

The general goal of this thesis is the identification and description of the geomorphological responses of the fluvial system to the Holocene Climate Changes, proposing a multi-sensor remote sensing approach. In particular, the specific aim of this work is the improvement of the present knowledge on the Holocene and historical morphodynamics of the Lower Mesopotamian waterscape, especially on the paleo-hydrology of the ancient Tigris-Euphrates fluvial system, focusing on the specific process in the dynamics of the waterscapes which plays a key role in the drainage network evolution in lowland areas. Crevasse splays represent significant geomorphological features for understanding the fluvial morphodynamics in lowland areas where avulsion processes prevail. The southern Mesopotamian Plain is the area where the ancient State of Lagash developed between the prehistoric Ubaid Period (c. 5200 - c. 3500 BC) and the late Parthian era (247 BC - AD 244), representing an ideal case study, where the Italian Archaeological Mission has been recently carried on extensive field-works at Tell Zurghul archaeological site. Here, an interdisciplinary approach, combining field surveys and geomorphological mapping through remote sensing techniques, has been applied for analyzing the function and role of the waterscape on the early civilization. Indeed, the geomorphological analysis through a remote sensing approach and the archaeological surveys are both essential for the reconstruction of a complex environmental system, where landforms due to different morphogenetic processes occur, related to the presence of a wide fluvial-deltaic paleo-system and early human societies. The main aim of the focus on this archaeological site is to contribute to the reconstruction of the surrounding waterscape and know more about waterscape-human interactions during the Holocene. The question of human-waterscape relationship worldwide has been and still is a central topic in geomorphological, environmental, and archaeological research. During the Holocene, the Tigris-Euphrates river system, in the lower sector of the Mesopotamian Plain (Iraq), has been characterized by complex morphodynamics in response to both climate fluctuations and extensive construction of artificial canals, dug since the first human settlements belonging to the Early River Valley Civilizations. The Lower Mesopotamian Plain (LMP) coincides with the southern Tigris and Euphrates deltaic plain, developed starting since the mid Holocene. During the early Holocene, the sea-level rise caused a general and rapid northward shifting of the Persian Gulf shoreline: the maximum marine ingression reached the area where the present towns of Nasiriyah and Al-Amara are located about 6000 yrs BP; after which the widespread progradation of the Tigris and Euphrates delta system accounted for the southward shoreline regression up to the present position. The development of a typical bird-foot delta guaranteed an amount of water indispensable for agriculture, cattle, settlements, and transport. Indeed, the high mobility of the channels and the frequent occurrence of avulsion processes (i.e., levees break and related crevasse splays formation) are the main features typically connected to a multi-channel system, guarantying the water supply through seasonal floods. In the area, the water management during the mid Holocene, digging an extensive network of canals and building several dams, can either improve the socio-economic conditions of a settlement or cause the end of another one. Within a wide floodplain characterized by very low elevation ranges such as the LMP, a remote sensing, multi-sensor approach is a suitable method for identifying the main geomorphological features related to the fluvial avulsion processes, describing the associated morphogenetic processes. Optical and multispectral Landsat 8 satellite images have been processed for computing NDVI and Clay Ratio indices, as well as to extract the Regions of Interest (ROIs) focused on the main features that made up a crevasse splay (i.e., crevasse channel, crevasse levee and crevasse deposit). The spectral signatures from active and abandoned crevasse splays have been extracted and compared among them, adopting four different methods of Supervised Classification. The analysis of the crevasse splays has been integrated with the investigation of the micro-topography leading to recognize the crevasse channels and levees, the upward convexity of the crevasse deposits and the distal or proximal position of the parent channel; the re-classification of different DEM sources, such as the optical AW3D30 and GDEM2 datasets with ground resolution of 1 arcsec (i.e., 30 m cell-1), leads to highlighting the “above-floodplain” topographic configuration of these landforms. The analysis here performed leads to investigating the entire Lower Mesopotamian Plain through both large and medium scale geomorphological investigation, identifying active and abandoned channels, discerning between active and abandoned avulsion processes and distinguishing crevasse channels, levees, and deposits. In like manner, human features are recognized, allowing the evaluation of human-environmental interactions.