All digital images are made up of thousands of dots of color, arranged in a grid. A pixel is the smallest part of that grid. So, a 1024-pixel wide image has 1024 dots in each row of dots. If the image is also 1024 pixels high, it has 1024 of these rows, making a little over 1 million dots, or pixels.
What you are doing by manipulating an image is changing the color or shade in many individual dots at the same time. If you type in 2000 into the zoom field and hit Enter, you can see these individual pixels in your texture image. If you hold the book at arms length and squint a little you'll see the colored pixels make up a nail in the wood.
The reason you resized to 1024 x 1024 pixels is because the image you had was too large (it held too much information) and so had a large file size. In gaming, we have to be constantly aware of the speed at which the game will render while playing it, so graphics file size is a big issue. In fact, 1024 isn't the final size you'll end up with. Later, you might decide to resize to 256 or 128 pixels wide. After all, a timber pallet is unlikely to be the showcase feature in your game level!
Textures should also be square if possible or a ratio of 2:1 or 1:2, such as 512 x 1024. For computer game graphics reasons, it is standard to make all graphics a power-of-two size. Get the number 2 and keep doubling it, and any of the sizes you get are ok (64, 128, 256, 512, 1024, 2048, 4096, and so on). This is traditionally because the computer memory allocates memory to textures in blocks of a power of two, so having a 520x520 image would still take up 1024x1024 of memory. Two of the more tangible benefits of this are that you work to standard sizes, cutting down the amount of decisions you have to make, and you can easily resize your texture to a range of file sizes by simply halving it a few times in GIMP.