Informations et ressources scientifiques
sur le développement des zones arides et semi-arides

Accueil du site → Doctorat → Allemagne → 2006 → Modeling water and salt dynamics under irrigated cotton with shallow groundwater in the Khorezm region of Uzbekistan

Rheinischen Friedrich-Wilhelms-Universität Bonn (2006)

Modeling water and salt dynamics under irrigated cotton with shallow groundwater in the Khorezm region of Uzbekistan

Forkutsa, I.

Titre : Modeling water and salt dynamics under irrigated cotton with shallow groundwater in the Khorezm region of Uzbekistan

Auteur  : Forkutsa, I.

Université de soutenance : Rheinischen Friedrich-Wilhelms-Universität Bonn

Grade : Doctoral Thesis 2006

Résumé In the Khorezm region of Uzbekistan, the main problems during cotton production remain poor management of irrigation water and the absence of efficient drainage, which causes a change in the local and regional hydrology and can lead to secondary soil salinization. This study aims at providing insights into how the different parts of the soil-plant-atmosphere system interact and where improvements could be implemented. Therefore, the soil water and salt balance of two cotton fields in the semi-arid region of Khorezm were examined. The one-dimensional, soil water model HYDRUS-1D based on the Richards equation was applied. During the 2002 and 2003 vegetation seasons, the research focused on three locations in a sandy loam field (field #1) and on two locations in a sandy field (field #2). For model calibration, the groundwater table and its salinity, the soil salinity (ECe), and the gravimetric soil water content and soil water pressure head was repeatedly determined across the survey fields. Using on-site meteorological measurements, the FAO-56 potential evapotranspiration (ET0) was calculated and separated into soil evaporation (E0) and transpiration (T0). The latter was fine-tuned with measurements on the development of the cotton leaf area index. Water applied during leaching and irrigation events was quantified. The simulated soil moisture data at four depths agreed well with the measured values (RMSE = 0.026 to 0.068 cm3 cm-3), whereas deviations between simulated and observed soil water potential could not be fully eliminated during model calibration (RMSE = 182 to 193 cm). Annual ET0 was estimated to be 994 mm in 2003. Simulated actual transpiration (Ta) was lower than potential transpiration (T0) due to the prevalence of water stress during the initial growth stage of cotton. On field #2, the simulated reduction in T was 45 % ( 100 mm), which was caused by a rapidly desiccating top-soil layer in combination with a shallow groundwater, resulting in poor cotton establishment under these waterlogged conditions. As expected, on field #1, the most severe water stress (81 % of T0) occurred at the location where least water was applied, highlighting the general problem related to a non-uniform water distribution along the field during irrigation. For all locations, the actual ET during the vegetation season 2003 was greater than the total amount of applied irrigation water. Overall, the soil was depleted by 32 mm of water on field #1 (0-90 cm soil depth) and by 60 mm on field #2 (0-28 cm soil depth). A capillary rise of 277 mm, 129 mm and 92 mm on field #1, beginning, mid and end, respectively, and 142 mm on field #2 in 2003 indicated that there was considerable groundwater contribution to ET. Moreover, the great heterogeneity of the results demonstrated that this contribution heavily depended upon the imposed irrigation management scheme. The simulated ECe for all locations was in a good agreement with the observed data at 20 cm soil depth. At 50, 80 and 105 cm depth, simulation results only poorly reflect the observed rapid salt fluctuations in 2003 (RMSE = 0.39 to 3.23 dS m-1), although the seasonal trend – desalination in response to leaching and slow re-salinization during the vegetation season – was correctly predicted by the model. Thus, the simulation results of solute transport provide reasonable insight into the seasonal salt balance of the monitored fields over the cotton root zone. At the beginning and middle of field #1, about 18 t salts per hectare were leached at the end of the simulation period 2003, while at the end of the field, 9 t of salts per hectare were leached due to the lowest amount of applied irrigation water. In field #2, no irrigation water was applied, and at the end of the simulated period there was a minor salt accumulation of 0.3 t per hectare. In general, the study allowed assessing the direct and indirect consequences of irrigation and basic agronomic mismanagement (field preparation and cultivation) on the soil water and salt regime. Furthermore, the model could precisely quantify the groundwater contribution and the (de-)salinization process. Based on the results, decisions can be made on how to improve the irrigation scheduling at the field level and how to prevent salt accumulation

Version intégrale

Page publiée le 1er janvier 2016, mise à jour le 31 décembre 2018