Temporal analysis of ramsar sites via remote sensing techniques - a case study of meke maar

Wetlands have an important place for human and natural life. Wetlands have globally significant social, economic and environmental benefits. Important functions of wetlands include water storage, regulation of groundwater level, protection from storms, control of flood risk, coastal stabilization, stabilization of local climatic conditions, mainly rainfall and temperature, erosion control, as well as carbon, nutrients, sediments and pollutants. Approximately 7% of the Earth's surface is covered by wetlands, while 40% of all species in the world and 12% of all animal species are home. Wetlands according to the Ramsar Convention; natural or artificial, continuous or seasonal, sweet, bitter or salty, permanent or flowing water bodies, marshes, peatlands, and sea water that do not exceed six meters in depth at the time of the tidal withdrawal. There are 14 Ramsar sites in Turkey and two of them are located in Konya Closed Basin which is one of the 25 basin located in Turkey. Meke Maarı, one of the Ramsar sites in Konya Kapalı Basin, is connected to the Karapınar district of Konya. Meke Maar took place on February 9, 2005 in the Wetlands of International Importance List. Meke Maar is a crater lake formed by filling a volcanic eruption pit 400 million years ago with water and islets in the middle. As a result of volcanic activity, bitter water contains magnesium and sodium sulphate, which causes low biodiversity. Aim of this study is Meke Maar, which is a Ramsar site, was examined from 1984 to 2017, along with the reasons for the change. Change analysis was conducted for 2 different month interval via remote sensing. Remote sensing can be defined as the collection of information remotely using sensors on objects without direct contact. Remote sensing knowledge involves the analysis and interpretation of electromagnetic spectrum measurements (EMR) observed or recorded at a point of observation by a sensor that is reflected or emitted from a target and is not in contact with an observer or target. The visible and near-infrared spectral bands in the satellite are used to distinguish specific surface features such as vegetation and water. The first month interval is May-June, while the second is August-September. LANDSAT satellite images were used for change analysis. Two different methods were used in the analysis for the may-june month interval. The first of these methods is the Normalized Difference Water Index (NDWI). Detection of water bodies from satellite images in a short period of time, the normalized difference between the two image bands is usually calculated, and then the thresholds to be determined are divided into two classes (water and non-water). NDWI is an improved method for describing open surface water properties and for detecting and measuring water presence on remote sensing images. NDWI utilizes reflected near-infrared radiation and visible green light to increase the presence of water features while removing the presence of soil and terrestrial vegetation features. The other method is Classification and Regression Tree (CART). Classification can be defined as the grouping of areas showing similarity in images by various mathematical methods according to determined criteria. CART includes the identification and organization of a binary decision tree using an educational set of knowledge of the correct classes, as in other supervised classification methods. CART can analyze complex data structures well in large data sets. The main idea is to separate the data set into homogeneous subgroups according to the same class. With this method, two classes were created on the image, namely water and other. The NDWI and CART methods were found to have the same trends of temporal change of May-June images. Therefore, only CART method was used for August-September. These methods were implemented in the Google Earth Engine (GEE) platform. GEE makes satellite images available for storage, layout and data mining on a global scale. The public data archive contains historical world images that are more than forty years old and new images are collected every day. GEE is a cloud-based platform that facilitates access to high-performance computing resources to handle very large geographic data sets. Thanks to this platform, satellite images can be analyzed on the platform without downloading. Earth Engine provides APIs in JavaScript and Python as well as other tools to enable analysis of large data sets. GEE has been found to facilitate the analysis of images. Accuracy assessment are done for NDWI and CART methods. Accuracy assessment refers to the correspondence of the results obtained from the satellite images to the real earth and is important in terms of the availability of the results. The most commonly used method to represent the degree of accuracy of a classification is to create k×k dimensional error matrix. The error matrices compare the relationship between the known reference data and the results of the automatic classification on a category basis. These two methods were found to be suitable for the detection of water areas and all of the results are greater than or equal to 90%. The water surface area of 2 months interval applied CART was not very different from each other and the average of water surface areas was taken. Water surface area was examined with meteorological parameters which are precipitation, evaporation and temperature. Parameters of 12 meteorological stations were used for Konya Closed Basin. According to these parameters, average precipitation of Konya Closed Basin is nearly half of the average precipitation of Turkey. 2 meteorological stations were used for Meke Maarı. These are Ereğli and Karapınar stations. Meteorological parameter of these stations were weighted according to distances between Meke Maar. Karapınar station is closer than Ereğli station. When Ereğli station was multiplied with 0.2, Karapınar station is was multiplied with 0.2. When the average water surface area obtained for 1984-2017 was examined, it was seen that it was in approximately same trend with the precipitation data until 2005. After 2005, the amount of water continued to decrease despite the increase in precipitation. In this case, the temporal change of groundwater level measurements were examined. There are 14 well in Konya Closed Basin and groundwater level measurements are continuously made from these wells until 2009. It has been seen to show a continuous decrease. The well, which is named as Gülfet Yayla, is close to Meke Maar. From 1984 to 2009, groundwater level decreased by approximately 22.5 meters. When the cause of this change is investigated, it has been observed that the cultivation of agricultural products, which have a high net water requirement is observed to increase. These products are sugar beet, corn and sunflower. From 1991 to 2009, sugar beet has increased by 250% in Karapınar. Corn plantation in Karapınar was planted in the area of 880 decares in 1999 and reached to 214360 decares in 2017. Like corn, sunflower was first cultivated in 2004 and has only 70 decares of land and it has reached 96000 decares in 2017. Net water requirements of these products are not met by precipitation and are trying to be met by wells. This leads to an increase in the number of run-off wells and also accelerates the decline of groundwater level. When the water samples taken were examined, it was seen that the water quality of Meke Maar was in the hypertrophic level and not suitable for agriculture. After 2005, it was determined that the cause of the decline in Meke Maar was caused by the lowering of the groundwater level. The use of agricultural products with a high net water requirement constituted the basis for this. As a result of the study, it was determined that the factors affecting the water change in Meke Maar were precipitation and groundwater level.