Local climate zone classification of surface energy flux distribution within an urban area of a hot-humid tropical city

<p>Monsoons were traditionally considered to be land-based systems. Recent definitions of monsoons based on either the seasonal reversal of winds or the local summer precipitation accounting for more than 50% of the annual precipitation suggests that monsoon domains extend over oceanic regions as well. The concept of global monsoon combines all the monsoon domains into a single entity. Modern observations show that the variations in precipitation are nearly coherent across all the individual monsoon domains on decadal timescales. Using a transient simulation of the global climate over the last 22,000 years as well as reanalysis data of the modern climate, we have shown that tropical precipitation has different characteristics over land and ocean grids. This is due to the differences in the energetics of monsoon over land and ocean grids. With a lower thermal heat capacity, the net surface energy flux over land is negligible, whereas it is quite large over the ocean. In fact, the orbital scale variability of net energy flux into the atmosphere over the ocean is controlled by the surface energy flux. Another major difference between land and ocean grids of the global monsoon is in the vertical profile of the vertical pressure velocity. It is bottom-heavy over land and top-heavy over the ocean. This results in smaller vertical transport of moist static energy (which has a minimum in the lower troposphere) over land, and a larger vertical transport over the ocean. These differences between the land and ocean, suggest that the land and ocean grids should not be combined as is traditionally done. Global monsoon-land and global monsoon-ocean should be studied separately.</p>

Download Full-text

Feature Importance Analysis for Local Climate Zone Classification Using a Residual Convolutional Neural Network with Multi-Source Datasets

Remote Sensing ◽

10.3390/rs10101572 ◽

2018 ◽

Vol 10 (10) ◽

pp. 1572 ◽

Cited By ~ 20

Author(s):

Chunping Qiu ◽

Michael Schmitt ◽

Lichao Mou ◽

Pedram Ghamisi ◽

Xiao Zhu

Keyword(s):

Neural Network ◽

Land Use ◽

Convolutional Neural Network ◽

Large Scale ◽

Landsat 8 ◽

Climate Zone ◽

Local Climate ◽

Feature Importance ◽

Local Climate Zone ◽

Sentinel 2

Global Local Climate Zone (LCZ) maps, indicating urban structures and land use, are crucial for Urban Heat Island (UHI) studies and also as starting points to better understand the spatio-temporal dynamics of cities worldwide. However, reliable LCZ maps are not available on a global scale, hindering scientific progress across a range of disciplines that study the functionality of sustainable cities. As a first step towards large-scale LCZ mapping, this paper tries to provide guidance about data/feature choice. To this end, we evaluate the spectral reflectance and spectral indices of the globally available Sentinel-2 and Landsat-8 imagery, as well as the Global Urban Footprint (GUF) dataset, the OpenStreetMap layers buildings and land use and the Visible Infrared Imager Radiometer Suite (VIIRS)-based Nighttime Light (NTL) data, regarding their relevance for discriminating different Local Climate Zones (LCZs). Using a Residual convolutional neural Network (ResNet), a systematic analysis of feature importance is performed with a manually-labeled dataset containing nine cities located in Europe. Based on the investigation of the data and feature choice, we propose a framework to fully exploit the available datasets. The results show that GUF, OSM and NTL can contribute to the classification accuracy of some LCZs with relatively few samples, and it is suggested that Landsat-8 and Sentinel-2 spectral reflectances should be jointly used, for example in a majority voting manner, as proven by the improvement from the proposed framework, for large-scale LCZ mapping.

Download Full-text